KLUGMC03_038-060HR 10/18/06 1:01 PM Page 38

C H A P T E R

3
Mendelian Genetics

Gregor Johann Mendel, who

in 1866 put forward the major
postulates of transmission
genetics as a result of
experiments with the
garden pea.

CHAPTER CONCEPTS
Inheritance is governed by information
stored in discrete factors called genes.
Genes are transmitted from generation
to generation on vehicles called
chromosomes.
Chromosomes, which exist in pairs, provide
the basis of biparental inheritance.
During gamete formation, chromosomes
are distributed according to postulates first
described by Gregor Mendel, based on his
nineteenth-century research with the
garden pea.
Mendelian postulates prescribe that
homologous chromosomes segregate from
one another and assort independently with
other segregating homologs during gamete
formation.
Genetic ratios, expressed as probabilities,
are subject to chance deviation and may be
evaluated statistically.
The analysis of pedigrees allows predictions
involving the genetic nature of
human traits.

38
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 39
3.2 The Monohybrid Cross Reveals How One Trait Is Transmitted from Generation to Generation 39
A
lthough inheritance of biological traits has been recog-
nized for thousands of years, the first significant
insights into the mechanisms involved occurred about
135 years ago. In 1866, Gregor Johann Mendel published the
results of a series of experiments that would lay the founda-
tion for the formal discipline of genetics. Mendels work went
largely unnoticed until the turn of the century, but in the ensu-
ing years the concept of the gene as a distinct hereditary unit
was established. The ways in which genes, as members of
chromosomes, are transmitted to offspring and control traits
were clarified. Research continued unabated throughout the
twentieth centuryindeed, studies in genetics, most recently
at the molecular level, have remained continually at the fore-
front of biological research since the early 1900s.
When Mendel began his studies of inheritance using Pisum
sativum, the garden pea, chromosomes and the role and mech-
anism of meiosis were totally unknown. Nevertheless, he
determined that discrete units of inheritance exist and pre-
dicted their behavior during the formation of gametes. Subse-
quent investigators, with access to cytological data, were able
to relate their observations of chromosome behavior during
meiosis to Mendels principles of inheritance. Once this corre-
lation was made, Mendels postulates were accepted as the
basis for the study of what is known as transmission genetics.
How Do We Know?
In this chapter, we focus on how Mendel was able to derive the
essential postulates that explain inheritance. As you study this
topic, you should try to answer several fundamental questions:
1. How did Mendel know that unit factors existed as
fundamental genetic components if he could not directly
observe them?
2. How do we know that an organism expressing a dominant
trait is homozygous or heterozygous?
3. In genetic data, how do we know that deviation from
the expected ratio is due to chance rather than another
independent factor?
4. How do we know how a trait is inherited in humans?
3.1 Mendel Used a Model
Experimental Approach to Study
Patterns of Inheritance
Johann Mendel was born in 1822 to a peasant family in the
central European village of Heinzendorf. An excellent student
in high school, he studied philosophy for several years after-
ward, and in 1843 he was admitted to the Augustinian
Monastery of St. Thomas in Brno, now part of the Czech
Republic, taking the name Gregor. In 1849, he was relieved of
pastoral duties and accepted a teaching appointment that last-
ed several years. From 1851 to 1853, he attended the Univer-
sity of Vienna, where he studied physics and botany. He
returned to Brno in 1854, where he taught physics and natural
science for the next 16 years. Mendel received support from
the monastery for his studies and research throughout his life.
In 1856, Mendel performed his first set of hybridization
experiments with the garden pea. The research phase of his
career lasted until 1868, when he was elected abbot of the
monastery. Although he retained his interest in genetics, his
new responsibilities demanded most of his time. In 1884,
Mendel died of a kidney disorder. The local newspaper paid
him the following tribute:
His death deprives the poor of a benefactor, and
mankind at large of a man of the noblest character, one
who was a warm friend, a promoter of the natural
sciences, and an exemplary priest.
Mendel first reported the results of some simple genetic
crosses between certain strains of the garden pea in 1865.
Although his was not the first attempt to provide experimental
evidence pertaining to inheritance, Mendels success where
others failed can be attributed, at least in part, to his elegant
model of experimental design and analysis.
Mendel showed remarkable insight into the methodology nec-
essary for good experimental biology. He chose an organism
that is easy to grow and hybridize artificially. The pea plant is
self-fertilizing in nature but is easy to crossbreed experimental-
ly. It reproduces well and grows to maturity in a single season.
Mendel followed seven visible features (unit characters), each
represented by two contrasting forms, or traits (Figure 31).
For the character stem height, for example, he experimented
with the traits tall and dwarf. He selected six other visibly con-
trasting pairs of traits involving seed shape and color, pod shape
and color, and pod and flower arrangement. From local seed
merchants, Mendel obtained true-breeding strainsthose in
which each trait appeared unchanged generation after generation
in self-fertilizing plants.
There were several reasons for Mendels success. In addi-
tion to his choice of a suitable organism, he restricted his
examination to one or very few pairs of contrasting traits in
each experiment. He also kept accurate quantitative records, a
necessity in genetic experiments. From the analysis of
his data, Mendel derived certain postulates that became
principles of transmission genetics.
The results of Mendels experiments were unappreciated until
the turn of the century, well after his death. However, once
Mendels publications were rediscovered by geneticists investi-
gating the function and behavior of chromosomes, the implica-
tions of his postulates were immediately apparent. He had
discovered the basis for the transmission of hereditary traits!
3.2 The Monohybrid Cross Reveals
How One Trait Is Transmitted
from Generation to Generation
Mendels simplest crosses involved only one pair of contrast-
ing traits. Each such experiment is a monohybrid cross,
which is made by mating true-breeding individuals from two
parent strains, each exhibiting one of the two contrasting forms
of the character under study. Initially, we examine the first
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 40

40 Chapter 3 Mendelian Genetics

Character Contrasting traits F1 results F2 results F2 ratio

5474 round
round/wrinkled all round 2.96:1
1850 wrinkled
Seeds
6022 yellow
yellow/green all yellow 3.01:1
2001 green

882 full
full/constricted all full 2.95:1
299 constricted
Pods
428 green
green/yellow all green 2.82:1
152 yellow

Flower violet/white 705 violet

color all violet 3.15:1
224 white

Flower 651 axial

position axial/terminal all axial 3.14:1
207 terminal

Stem 787 tall

tall/dwarf all tall 2.84:1
length 277 dwarf

FIGURE 3-1 Seven pairs of contrasting traits and the results of Mendels seven monohybrid crosses of the garden pea (Pisum sativum).
In each case, pollen derived from plants exhibiting one trait was used to fertilize the ova of plants exhibiting the other trait. In the F1
generation, one of the two traits was exhibited by all plants. The contrasting trait reappeared in approximately 1>4 of the F2 plants.

generation of offspring of such a cross, and then we consider
the results of selfing, the offspring of self-fertilizing individu-
als from this first generation. The original parents constitute
the P1 , or parental, generation, their offspring are the F1 , or
first filial generation, and the individuals resulting from the
selfed F1 generation are the F2 , or second filial generation.
We can continue to follow subsequent generations.
The cross between true-breeding pea plants with tall stems
and dwarf stems is representative of Mendels monohybrid
crosses. Tall and dwarf are contrasting traits of the character
of stem height. Unless tall or dwarf plants are crossed togeth-
er or with another strain, they will undergo self-fertilization
and breed true, producing their respective traits generation
after generation. However, when Mendel crossed tall plants
with dwarf plants, the resulting F1 generation consisted only
of tall plants. When members of the F1 generation were
selfed, Mendel observed that 787 of 1064 F2 plants were tall,
while the remaining 277 were dwarf. Note that in this cross
(Figure 31) the dwarf trait disappears in the F1 , only to
reappear in the F2 generation.
Genetic data are usually expressed and analyzed as ratios. In
this particular example, many identical P1 crosses were made,
and many F1 plantsall tallwere produced. Of the 1064 F2
offspring, 787 were tall and 277 were dwarfa ratio of
2.84:1.0, or about 3:1.
Mendel made similar crosses between pea plants exhibiting
other pairs of contrasting traits; the results of these crosses are
shown in Figure 31. In every case, the outcome was similar
to the tall/dwarf cross just described. All F1 offspring were
identical to one of the parents, but in the F2 offspring, an
approximate ratio of 3:1 was obtained. That is, three-fourths
looked like the F1 plants, while one-fourth exhibited the con-
trasting trait, which had disappeared in the F1 generation.
We will point out one further aspect of Mendels monohy-
brid crosses. In each cross, the F1 and F2 patterns of inheri-
tance were similar regardless of which P1 plant served as the
source of pollen (sperm) and which served as the source of the
ovum (egg). The crosses could be made either waypollination
of dwarf plants by tall plants or vice versa. These are called
reciprocal crosses. Therefore, the results of Mendels mono-
hybrid crosses were not sex-dependent.
To explain these results, Mendel proposed the existence of
particular unit factors for each trait. He suggested that these
factors serve as the basic units of heredity and are passed
unchanged from generation to generation, determining the var-
ious traits expressed by each individual plant. Using these gen-
eral ideas, Mendel proceeded to hypothesize precisely how unit
factors could account for the results of the monohybrid crosses.
Mendels First Three Postulates
Using the consistent pattern of results in the monohybrid
crosses, Mendel derived the following three postulates or
principles of inheritance.
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 41
3.2 The Monohybrid Cross Reveals How One Trait Is Transmitted from Generation to Generation 41
1. Unit Factors in Pairs
Genetic characters are controlled by unit factors that
exist in pairs in individual organisms.
In the monohybrid cross involving tall and dwarf stems, a
specific unit factor exists for each trait. Because the fac-
tors occur in pairs, three combinations are possible: two
factors for tallness, two factors for dwarfness, or one fac-
tor for each trait. Every individual contains one of these
three combinations, which determines stem height.
2. Dominance/Recessiveness
When two unlike unit factors responsible for a single char-
acter are present in a single individual, one unit factor is
dominant to the other, which is said to be recessive.
In each monohybrid cross, the trait expressed in the F1
generation is controlled by the dominant unit factor. The
trait not expressed is controlled by the recessive unit fac-
tor. Note that this dominance/recessiveness relationship
pertains only when unlike unit factors are present in
pairs. The terms dominant and recessive are also used to
designate traits. In this case, tall stems are said to be
dominant over the recessive dwarf stems.
3. Segregation
During the formation of gametes, the paired unit factors
separate or segregate randomly so that each gamete
receives one or the other with equal likelihood.
If an individual contains a pair of like unit factors
(e.g., both specific for tall), then all gametes receive one
tall unit factor. If an individual contains unlike unit fac-
tors (e.g., one for tall and one for dwarf), then each
gamete has a 50 percent probability of receiving either
the tall or the dwarf unit factor.
These postulates provide a suitable explanation for the
results of the monohybrid crosses. Lets use the tall/dwarf
cross to illustrate. Mendel reasoned that P1 tall plants contain
identical paired unit factors, as do the P1 dwarf plants. The
gametes of tall plants all receive one tall unit factor as a result
of segregation. Similarly, the gametes of dwarf plants all
receive one dwarf unit factor. Following fertilization, all F1
plants receive one unit factor from each parent: a tall factor
from one and a dwarf factor from the other, reestablishing the
paired relationshipbut because tall is dominant to dwarf, all
F1 plants are tall.
When F1 plants form gametes, the postulate of segregation
demands that each gamete randomly receives either the tall or
the dwarf unit factor. Following random fertilization events dur-
ing F1 selfing, four F2 combinations result in equal frequency:
1. tall/tall
2. tall/dwarf
3. dwarf/tall
4. dwarf/dwarf
Combinations (1) and (4) result in tall and dwarf plants,
respectively. According to the postulate of dominance/
recessiveness, combinations (2) and (3) both yield tall plants.
Therefore, the F2 is predicted to consist of 3>4 tall and 1>4
dwarf, or a ratio of 3:1. This is approximately what Mendel
observed in the cross between tall and dwarf plants. A similar
pattern was observed in each of the other monohybrid crosses
(see Figure 31).
Modern Genetic Terminology
To illustrate the monohybrid cross and Mendels first three
postulates, we must first introduce several new terms as well
as a symbol convention for the unit factors.
Traits such as tall or dwarf are visible expressions of the
information contained in unit factors. The physical appear-
ance of a trait is the phenotype of the individual. Mendels
unit factors represent units of inheritance called genes by
modern geneticists. For any given character, such as plant
height, the phenotype is determined by alternative forms of a
single gene called alleles. For example, the unit factors repre-
senting tall and dwarf are alleles determining the height of the
pea plant.
The convention we will use is to choose the first letter of
the recessive trait to symbolize the character in question
the lowercase italic letter designates the allele for the reces-
sive trait, and the uppercase italic letter designates the allele
for the dominant trait. Thus, for Mendels pea plants, we
use d for the dwarf allele and D for the tall allele. When
alleles are written in pairs to represent the two unit factors
present in any individual (DD, Dd, or dd), these symbols
are called the genotype. This term reflects the genetic
makeup of an individual, whether it is haploid or diploid.
By reading the genotype, we know the phenotype of the
individual: DD and Dd are tall, and dd is dwarf. When both
alleles are the same (DD or dd), the individual is
homozygous or a homozygote; when the alleles are differ-
ent (Dd), we use the term heterozygous or a heterozygote.
These symbols and terms are used in Figure 32 to illus-
trate the monohybrid cross.
Because he operated without the hindsight that modern
geneticists enjoy, Mendels analytical reasoning must be
considered a truly outstanding scientific achievement. On the
basis of rather simple but precisely executed breeding experi-
ments, he not only proposed that discrete particulate units of
heredity exist, he also explained how they are transmitted from
one generation to the next.
Now Solve This
Problem 5 on page 58 involves a Mendelian cross where
you must determine the mode of inheritance and the
genotypes of the parents in a number of instances.
Hint: The first step is to determine how many genes
are involved. To do so, convert the data to ratios that
are characteristic of Mendelian crosses. In the case of
this problem, ask first whether any of the F2 ratios
match Mendels 3:1 monohybrid ratio.
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 42

42 Chapter 3 Mendelian Genetics

Punnett Squares our F1 * F1 monohybrid cross. Each of the possible

The genotypes and phenotypes resulting from the recombi-
nation of gametes during fertilization can be easily visual-
ized by constructing a Punnett square, named after the
person who first devised this approach, Reginald C. Pun-
nett. Figure 33 demonstrates this method of analysis for
P1 cross
gametes is assigned to a column or a row; the vertical col-
umn represents those of the female parent, and the horizon-
tal row represents those of the male parent. After putting
the gametes into the rows and columns, the new generation
is predicted by combining the male and female gametic
information for each combination by entering the resulting
genotypes in the boxes. This process thus lists all possible

Phenotypes: tall dwarf F1 cross

Dd
Dd
Genotypes: DD
dd
tall tall

Gamete formation
Gamete formation
by F1 generation
DD dd
Dd Dd
D Gametes d

D d D d

F1 generation

D d Setting up a
Fertilization Punnett square

D d
Dd
all tall D

F1 cross

Dd
Dd Filling out squares

representing fertilization

D d
D d D d
D DD Dd
F1 gametes tall tall
d dD dd
Monohybrid Crosses and Punnett Squares

tall dwarf

F2 generation

F1 gametes D d
D d F2 results

Random fertilization Genotype Phenotype

1 DD
Web Tutorial 3.1

3/4 tall
F2 genotypes DD Dd Dd dd 2 Dd
F2 phenotypes tall tall tall dwarf
1 dd 1/4 dwarf
Designation Homozygous Heterozygous Heterozygous Homozygous
1:2:1 3:1

FIGURE 3-2 The monohybrid cross between tall (D) and dwarf FIGURE 3-3 A Punnett square generating the F2
(d) pea plants. Individuals are shown in rectangles, and gametes in circles. ratio of the F1 * F1 cross shown in Figure 32.
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 43

3.3 Mendels Dihybrid Cross Generated a Unique F2 Ratio 43

random fertilization events. The genotypes and phenotypes Testcross results

of all potential offspring are ascertained by reading the (a) (b)
entries in the boxes.
DD
dd Dd
dd
The Punnett square method is particularly useful when you
are first learning about genetics and how to solve problems. Homozygous Homozygous Heterozygous Homozygous
tall dwarf tall dwarf
Note the ease with which the 3:1 phenotypic ratio and
the 1:2:1 genotypic ratio is derived in the F2 generation in
Figure 33. D d D d d

The Testcross: One Character

Tall plants produced in the F 2 generation are predicted to
be either the DD or Dd genotype. You might ask if there is
a way to distinguish the genotype. Mendel devised a
rather simple method that is still used today in breeding
plants and animals: the testcross. The organism express-
ing the dominant phenotype, but of unknown genotype, is
crossed to a known homozygous recessive individual. For
example, as shown in Figure 34(a) , if a tall plant of
genotype DD is testcrossed to a dwarf plant, which must
have the dd genotype, all offspring will be tall phenotypi-
cally and Dd genotypically. However, as shown in Figure
34(b) , if a tall plant is Dd and it is crossed to a dwarf
plant (dd), then one-half of the offspring will be tall (Dd)
and the other half will be dwarf (dd). Therefore, a 1:1
tall/dwarf ratio demonstrates the heterozygous nature of
the tall plant of unknown genotype. The test cross rein-
forced Mendels conclusion that separate unit factors
control traits.
Dd Dd dd
all tall 1/2 tall 1/2 dwarf
FIGURE 3-4 Test cross of a single character. In (a), the tall parent is
homozygous, but in (b), the tall parent is heterozygous. The genotype
of each tall P1 plant can be determined by examining the offspring
when each is crossed to a homozygous recessive dwarf plant.
3.3 Mendels Dihybrid Cross
Generated a Unique F2 Ratio
As a natural extension of the monohybrid cross, Mendel also
designed experiments in which he examined two characters
simultaneously. Such a cross, involving two pairs of contrasting
traits, is a dihybrid cross, or two-factor cross. For example, if
pea plants having yellow seeds that are round are bred with
those having green seeds that are wrinkled, the results shown in

How Mendels Peas Become Wrinkled: A Molecular Explanation

O nly recently, well over a hundred osmotic balance that minimizes the loss of coding sequence. This foreign segment
years after Mendel used wrinkled water. The end result is a smooth-textured closely resembles other such sequences,
peas in his groundbreaking hybridization outer coat. called transposable elements. These
experiments, have we come to find out The SBEI gene has been cloned and ana- sequences have the ability to move from
how the wrinkled gene makes peas wrin- lyzed, providing greater insight into the place to place in the genome of organ-
kled. The wild-type allele of the gene relationship between genotypes and isms. Transposable elements have been
encodes a protein called starch-branch- phenotypes. Interestingly, the mutant found in maize (corn), parsley, and snap-
ing enzyme (SBEI). This enzyme cat- gene contains a foreign sequence of some dragons, fruit flies, and humans, among
alyzes the formation of highly branched 800 base pairs that disrupts the normal many other organisms.
starch molecules as the seed matures.
Wrinkled peas, which result from the
homozygous presence of the mutant form
of the gene, lack the activity of this
enzyme. The production of branch points
is inhibited during the synthesis of starch
within the seed, which in turn leads to the
accumulation of more sucrose and a high-
er water content while the seed develops.
Osmotic pressure inside rises, causing the
seed to lose water internally, and ultimate-
ly results in the wrinkled appearance of
the seed at maturation. In contrast, devel-
oping seeds that bear at least one copy of
the normal gene (being either homozy-
gous or heterozygous for the dominant Wrinkled and round garden peas, the phenotypic traits in one of Mendels monohybrid
allele) synthesize starch and reach an crosses.
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 44

44 Chapter 3 Mendelian Genetics

P1 cross P1 cross inherited independently of each other;

yellow, round
green, wrinkled yellow, wrinkled
green, round
F1
All yellow, round
F1
F1 yellow, round
yellow, round
F2 9/16 yellow, round 3/16 green, round
3/16 yellow, wrinkled 1/16 green, wrinkled
FIGURE 3-5 F1 and F2 results of Mendels dihybrid crosses, where
the plants on the top left with yellow, round seeds are crossed with
plants having green, wrinkled seeds, and the plants on the top right
with yellow, wrinkled seeds are crossed with plants having green,
round seeds.
Figure 35 will occur: the F1 offspring will all be yellow and
round. It is therefore apparent that yellow is dominant to green
and that round is dominant to wrinkled. When the F1 individu-
als are selfed, approximately 9>16 of the F2 plants express yel-
low and round, 3>16 express yellow and wrinkled, 3>16
express green and round, and 1>16 express green and wrinkled.
A variation of this cross is also shown in Figure 35. Instead
of crossing one P1 parent with both dominant traits (yellow,
round) and one with both recessive traits (green, wrinkled),
plants with yellow, wrinkled seeds are crossed with those with
green, round seeds. In spite of the change in the P1 pheno-
types, both the F1 and F2 results remain unchanged. It will
become clear in the next section why this is so.
Mendels Fourth Postulate:
Independent Assortment
We can most easily understand the results of a dihybrid cross if
we consider it theoretically as consisting of two monohybrid
crosses conducted separately. Think of the two sets of traits as
that is, the chance of any plant having
yellow or green seeds is not at all influ-
enced by the chance that this plant will
have round or wrinkled seeds. Thus,
because yellow is dominant to green, all
F1 plants in the first theoretical cross
would have yellow seeds. In the second
theoretical cross, all F1 plants would have
round seeds because round is dominant to
wrinkled. When Mendel examined the F1
plants of the dihybrid cross, all were yel-
low and round, as we just predicted.
The predicted F2 results of the first
cross are 3/4 yellow and 1/4 green. Sim-
ilarly, the second cross would yield 3/4
round and 1/4 wrinkled. Figure 35
shows that in the dihybrid cross,
12/16 F2 plants are yellow while 4/16 are
green, exhibiting the expected 3:1 (3/4:1/4) ratio. Similarly,
12/16 F2 plants have round seeds while 4/16 have wrinkled
seeds, again revealing the 3:1 ratio.
It is evident that the two pairs of contrasting traits are inher-
ited independently, so we can predict the frequencies of all
possible F2 phenotypes by applying the product law of prob-
abilities: When two independent events occur simultaneously,
the combined probability of the two outcomes is equal to the
product of their individual probabilities of occurrence. For
example, the probability of an F2 plant having yellow and
round seeds is (3/4)(3/4), or 9/16, because 3/4 of all F2 plants
should be yellow and (3/4) of all F2 plants should be round. In
a like manner, the probabilities of the other three F2 pheno-
types can be calculated: yellow (3/4) and wrinkled (1/4) are
predicted to be present together 3/16 of the time; green (1/4)
and round (3/4) are predicted 3/16 of the time; and green (1/4)
and wrinkled (1/4) are predicted 1/16 of the time. These cal-
culations are shown in Figure 36.
It is now apparent why the F1 and F2 results are identical
whether the initial cross is yellow, round plants bred with
green, wrinkled plants, or if yellow, wrinkled plants are bred
with green, round plants. In both crosses, the F1 genotype of
all plants is identical. Each plant is heterozygous for both
gene pairs. As a result, the F2 generation is also identical in
both crosses.

FIGURE 3-6 Computation of the

F1 yellow, round yellow, round
combined probabilities of each F2
F2 Of all offspring Of all offspring Combined probabilities phenotype for two independently
Independent Assortment

inherited characters. The

probability of each plants being
Web Tutorial 3.2

3/4 are round (3/4)(3/4) = 9/16 yellow, round yellow or green is independent of
3/4 are yellow and the probability of its bearing round
1/4 are wrinkled (3/4)(1/4) = 3/16 yellow, wrinkled or wrinkled seeds.

3/4 are round (1/4)(3/4) = 3/16 green, round

1/4 are green and
1/4 are wrinkled (1/4)(1/4) = 1/16 green, wrinkled
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 45

3.3 Mendels Dihybrid Cross Generated a Unique F2 Ratio 45

On the basis of similar results in numerous dihybrid crosses,
Mendel proposed a fourth postulate called independent
assortment: During gamete formation, segregating pairs of
unit factors assort independently of each other.
This postulate stipulates that segregation of any pair of unit
factors occurs independently of all others. As a result of
random segregation, each gamete receives one member of
every pair of unit factors. For one pair, whichever unit factor is
received does not influence the outcome of segregation of any
other pair. Thus, according to the postulate of independent
assortment, all possible combinations of gametes are formed
in equal frequency.
The Punnett square in Figure 37 shows how independent
assortment works in the formation of the F2 generation.

FIGURE 3-7 Analysis of the dihybrid P1 Cross P1 Cross

crosses shown in Figure 35. The F1
heterozygous plants are self-fertilized to GGWW
ggww GGww
ggWW
produce an F2 generation, which is
computed using a Punnett square. Both the yellow, round green, wrinkled yellow, wrinkled green, round
phenotypic and genotypic F2 ratios are Gamete Gamete
shown. formation formation
GW gw Gw gW
Fertilization Fertilization

GgWw

F1 yellow, round (in both cases)

F1 cross GgWw
GgWw

GW Gw gW gw

GW
GGWW GGWw GgWW GgWw
yellow, round yellow, round yellow, round yellow, round

Gw
GGWw GGww GgWw Ggww
yellow, round yellow, wrinkled yellow, round yellow, wrinkled

gW
GgWW GgWw ggWW ggWw
yellow, round yellow, round green, round green, round

gw
GgWw Ggww ggWw ggww
yellow, round yellow, wrinkled green, round green, wrinkled

F2 Generation

F2 Genotypic ratio F2 Phenotypic ratio

1/16 GGWW
2/16 GGWw 9/16 yellow, round
2/16 GgWW
4/16 GgWw

1/16 GGww
3/16 yellow, wrinkled
2/16 Ggww

1/16 ggWW
3/16 green, round
2/16 ggWw

1/16 ggww 1/16 green, wrinkled

KLUGMC03_038-060HR 10/18/06 1:01 PM Page 46
46 Chapter 3 Mendelian Genetics
Examine the formation of gametes by the F1 plants; segrega-
tion prescribes that every gamete receives either a G or g
allele and a W or w allele. Independent assortment stipulates
that all four combinations (GW, Gw, gW, and gw) will be
formed with equal probabilities.
In every F1 * F1 fertilization event, each zygote has an
equal probability of receiving one of the four combinations
from each parent. If many offspring are produced, 9/16 have
yellow, round seeds, 3/16 have yellow, wrinkled seeds, 3/16
have green, round seeds, and 1/16 have green, wrinkled seeds,
yielding what is designated as Mendels 9:3:3:1 dihybrid
ratio. This is an ideal ratio based on probability events involv-
ing segregation, independent assortment, and random fertil-
ization. Because of deviation due strictly to chance,
particularly if small numbers of offspring are produced, actu-
al results are highly unlikely to match the ideal ratio.
The Testcross: Two Characters
The testcross can also be applied to individuals that express
two dominant traits but whose genotypes are unknown. For
example, the expression of the yellow, round seed phenotype
in the F2 generation just described may result from the
GGWW, GGWw, GgWW, and GgWw genotypes. If an F2 yel-
low, round plant is crossed with a homozygous recessive
green, wrinkled plant (ggww), analysis of the offspring will
indicate the actual genotype of that yellow, round plant. Each
of the above genotypes results in a different set of gametes,
and in a testcross, a different set of phenotypes in the resulting
offspring. You should work out the results of each of these
four crosses to be sure you understand this concept.
Now Solve This
Problem 7 on page 59 involves a series of Mendelian
dihybrid crosses where you must determine the geno-
types of the parents in a number of instances.
Hint: In each case, write down everything that you
know for certain. This reduces the problem to its bare
essentials, clarifying what you need to determine. For
example, the wrinkled, yellow plant in case (b) must
be homozygous for the recessive wrinkled alleles and
bear at least one dominant allele for the yellow trait.
Having established this, you need only determine the
remaining allele for cotyledon color.
3.4 The Trihybrid Cross Demonstrates
That Mendels Principles Apply to
Inheritance of Multiple Traits
Thus far, we have considered inheritance by individuals of up
to two pairs of contrasting traits. Mendel demonstrated that
the identical processes of segregation and independent assort-
ment apply to three pairs of contrasting traits in what is called
a trihybrid cross, or three-factor cross.
P1 AABBCC
aabbcc
Gametes ABC abc
F1 AaBbCc
ABC ABc AbC Abc
Gametes
aBC aBc abC abc
FIGURE 3-8 Formation of P1 and F1 gametes in a trihybrid cross.
Although a trihybrid cross is somewhat more complex than
a dihybrid cross, its results are easily calculated if the princi-
ples of segregation and independent assortment are followed.
For example, consider the cross shown in Figure 38, where
the gene pairs of theoretical contrasting traits are represented
by the symbols A, a, B, b, C, and c. In the cross between
AABBCC and aabbcc individuals, all F1 individuals are het-
erozygous for all three gene pairs. Their genotype, AaBbCc,
results in the phenotypic expression of the dominant A, B, and
C traits. When F1 individuals serve as parents, each produces
eight different gametes in equal frequencies. At this point, we
could construct a Punnett square with 64 separate boxes and
read out the phenotypesbut such a method is cumbersome
in a cross involving so many factors. Therefore another
method has been devised to calculate the predicted ratio.
The Forked-Line Method
It is much less difficult to consider each contrasting pair of
traits separately and then to combine these results by using
the forked-line method, first shown in Figure 36. This
method, also called a branch diagram, relies on the simple
application of the laws of probability established for the
dihybrid cross. Each gene pair is assumed to behave indepen-
dently during gamete formation.
When the monohybrid cross AA * aa is made, we know
that:
1. All F1 individuals have the genotype Aa and express the
phenotype represented by the A allele, which is called the
A phenotype in the following discussion.
2. The F2 generation consists of individuals with either the
A phenotype or the a phenotype in the ratio of 3:1.
The same generalizations can be made for the BB * bb
and CC * cc crosses. Thus, in the F2 generation, 3/4 of all
organisms express phenotype A, 3/4 express B, and 3/4
express C. Similarly, 1/4 of all organisms express phenotype
a, 1/4 express b, and 1/4 express c. The proportions of organ-
isms that express each phenotypic combination can be pre-
dicted by assuming that fertilization, following the
independent assortment of these three gene pairs during
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 47

3.5 Mendels Work Was Rediscovered in the Early Twentieth Century 47

Generation of F2 trihybrid phenotypes
The source of natural variation
A or a B or b C or c Combined proportion intrigued students of evolutionary theo-
3/4 C (3/4)(3/4)(3/4) ABC = 27/64 ABC
ry. These individuals, stimulated by the
3/4 B proposal developed by Charles Darwin
1/4 c (3/4)(3/4)(1/4) ABc = 9/64 ABc and Alfred Russel Wallace, believed in
3/4 A
3/4 C (3/4)(1/4)(3/4) AbC = 9/64 AbC continuous variation, whereby off-
1/4 b spring were a blend of their parents
1/4 c (3/4)(1/4)(1/4) Abc = 3/64 Abc
phenotypes. As we mentioned earlier,
3/4 C (1/4)(3/4)(3/4) aBC = 9/64 aBC Mendel theorized that variation was
3/4 B
1/4 c (1/4)(3/4)(1/4) aBc = 3/64 aBc due to discrete or particulate units,
1/4 a resulting in discontinuous variation.
3/4 C (1/4)(1/4)(3/4) abC = 3/64 abC
1/4 b For example, Mendel proposed that the
1/4 c (1/4)(1/4)(1/4) abc = 1/64 abc F2 offspring of a dihybrid cross are
expressing traits produced by new com-
FIGURE 3-9 Generation of the F2 trihybrid phenotypic ratio, using binations of previously existing unit factors. As a result,
the forked-line method. This method is based on the expected Mendels theories did not fit well with the evolutionists pre-
probability of occurrence of each phenotype. conceptions about causes of variation.
In the latter part of the nineteenth century, a remarkable
observation set the scene for the rebirth of Mendels work:
gamete formation, is a random processwe simply apply the
Walter Flemmings discovery of chromosomes in the nuclei of
product law of probabilities once again. Figure 39 uses the
salamander cells. In 1879, Flemming described the behavior of
forked-line method to calculate the phenotypic proportions of
these threadlike structures during cell division. As a result of
the F2 generation. They fall into the trihybrid ratio of
his findings and the work of many other cytologists, the pres-
27:9:9:9:3:3:3:1. The same method can be used to solve
ence of discrete units within the nucleus soon became an inte-
crosses involving any number of gene pairs, provided that all
gral part of ideas about inheritance. It was this mind-set that
gene pairs assort independently of each other. We shall see
prompted scientists to reexamine Mendels findings.
later that this is not always the case. However, it appeared to
In the early twentieth century, research led to renewed inter-
be true for all of Mendels characters.
est in Mendels work. Hybridization experiments similar to
Mendels were performed independently by three botanists:
Hugo de Vries, Karl Correns, and Erich Tschermak. De
Vriess work demonstrated the principle of segregation in
Now Solve This his experiments with several plant species. Apparently, he
Problem 15 on page 59 asks you to use the forked-line searched the existing literature and found that Mendels work
method to determine the outcome of a number of tri- anticipated his own conclusions! Correns and Tschermak also
hybrid crosses. reached conclusions similar to those of Mendel.
In 1902, two cytologists, Walter Sutton and Theodor Boveri,
Hint: In using the forked-line method, consider each
gene pair separately. For example, in this problem,
independently published papers linking their discoveries of
first predict the outcome of each cross for A>a genes, the behavior of chromosomes during meiosis to the
then for the B>b genes, and finally, for the C>c genes. Mendelian principles of segregation and independent assort-
Then you are prepared to pursue the outcome of each ment. They pointed out that the separation of chromosomes
cross using the forked-line method. during meiosis could serve as the cytological basis of these
two postulates. Although they thought Mendels unit factors
were probably chromosomes rather than genes on chromo-
somes, their findings also reestablished the importance of
Mendels work, which became the basis of ensuing genetic
3.5 Mendels Work Was Rediscovered
investigations. Sutton and Boveri are credited with initiating
in the Early Twentieth Century the chromosomal theory of inheritance, which was devel-
Mendels work, initiated in 1856, was presented to the Brnn oped during the next two decades.
Society of Natural Science in 1865 and published the follow-
ing year. While his findings were often cited and discussed, Unit Factors, Genes,
their significance went unappreciated for about 35 years. and Homologous Chromosomes
Many reasons have been suggested to explain why the signif- Because the correlation between Suttons and Boveris obser-
icance of his research was not immediately recognized. vations and Mendelian principles serves as the foundation for
Mendels adherence to mathematical analysis of probability the modern interpretation of transmission genetics, we will
events was an unusual approach in those days for biological examine this correlation in some depth before moving on to
studies. Perhaps his approach seemed foreign to his contempo- other topics.
raries. More important, his conclusions did not fit well with As we know, each species possesses a specific number of
existing theories on the cause of variation among organisms. chromosomes in each somatic cell nucleus (except in
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 48

48 Chapter 3 Mendelian Genetics

gametes). For diploid organisms, this number is called the
diploid number (2n) and is characteristic of that species. Dur-
ing the formation of gametes, this number is precisely halved
(n), and when two gametes combine during fertilization, the
diploid number is reestablished. During meiosis, however, the
chromosome number is not reduced in a random manner. It
was apparent to early cytologists that the diploid number of
chromosomes is composed of homologous pairs identifiable
by their morphological appearance and behavior. The gametes
contain one member of each pairthus the chromosome com-
plement of a gamete is quite specific, and the number of chro-
mosomes in each gamete is equal to the haploid number.
With this basic information, we can see the correlation
between the behavior of unit factors and chromosomes
and genes. Figure 310 shows three of Mendels postu-
lates and the chromosomal explanation of each. Unit fac-
tors are really genes located on homologous pairs of
chromosomes [Figure 310(a)]. Members of each pair of
homologs separate, or segregate, during gamete formation
[Figure 310(b)]. Two different alignments are possible,
both of which are shown.
To illustrate the principle of independent assortment, we
must distinguish between members of any given homologous
pair of chromosomes. One member of each pair comes from

FIGURE 3-10 Illustrated

(a) Unit factors in pairs (first meiotic prophase)
correlation between the
Mendelian postulates of
Homologous chromosomes in pairs (a) unit factors in pairs,
G G g g (b) segregation, and
(c) independent
assortment, showing the
presence of genes located
on homologous
W W w w chromosomes and their
behavior during meiosis.
Genes are part of chromosomes

(b) Segregation of unit factors during gamete formation (first meiotic anaphase)

Homologs segregate during meiosis

G g G g

G g G g

or

W w w W
W w w W
Each pair separates Each pair separates

(c) Independent assortment of segregating unit factors (following many meiotic events)

Nonhomologous chromosomes assort independently

G g G g

W w w W

1/4 1/4 1/4 1/4

All possible gametic combinations are formed with equal probablility

KLUGMC03_038-060HR 10/18/06 1:01 PM Page 49
the maternal parent, while the other member comes from the
paternal parent. (We represent the different parental origins
with different colors.) As shown in Figure 310(c), following
independent segregation of each pair of homologs, each
gamete receives one member from each pair of chromosomes.
All possible combinations are formed with equal probability.
If we add the symbols used in Mendels dihybrid cross
(G, g and W, w) to the diagram, we can see why equal num-
bers of the four types of gametes are formed. The independent
behavior of Mendels pairs of unit factors (G and W in this
example) is due to their presence on separate pairs of homol-
ogous chromosomes.
Observations of the phenotypic diversity of living organ-
isms make it logical to assume that there are many more
genes than chromosomes. Therefore, each homolog must
carry genetic information for more than one trait. The cur-
rently accepted concept is that a chromosome is composed
of a large number of linearly ordered, information-containing
genes. Mendels unit factors (which determine tall or dwarf
stems, for example) actually constitute a pair of genes locat-
ed on one pair of homologous chromosomes. The location
on a given chromosome where any particular gene occurs is
called its locus (pl., loci). The different forms taken by a
given gene, called alleles (G or g), contain slightly different
genetic information that determines the same character
(seed color in this case). Although we have examined only
genes with two alternative alleles, most genes have more
than two allelic forms. We conclude this section by review-
ing the criteria necessary to classify two chromosomes as a
homologous pair:
1. During mitosis and meiosis, when chromosomes are vis-
ible as distinct figures, both members of a homologous
pair are the same size and exhibit identical centromere
locations. The sex chromosomes are an exception.
2. During early stages of meiosis, homologous chromosomes
form pairs, or synapse.
3. Although not generally microscopically visible, homologs
contain identical, linearly ordered gene loci.
3.6 Independent Assortment Leads
to Extensive Genetic Variation
One major consequence of independent assortment is the pro-
duction by an individual of genetically dissimilar gametes.
Genetic variation results because the two members of any
homologous pair of chromosomes are rarely, if ever, geneti-
cally identical. Therefore, because independent assortment
leads to the production of all possible chromosome combina-
tions, extensive genetic diversity results.
We have seen that the number of possible gametes, each
with different chromosome compositions, is 2n , where n
equals the haploid number. Thus, if a species has haploid
number of n = 4, then 24 = 16 different gamete combina-
tions can be formed as a result of independent assortment.
Although this number is not high, consider the human
3.7 Laws of Probability Help to Explain Genetic Events 49
species, where n = 23. If 223 is calculated, we find that in
excess of 8 * 106 , or over 8 million, different types of
gametes are represented. Because fertilization represents an
event involving only one of approximately 8 * 106 possible
gametes from each of two parents, each offspring represents
only one of (8 * 106)2 , or one of only 64 * 1012 potential
genetic combinations! No wonder that, except for identical
twins, each member of the human species demonstrates a dis-
tinctive appearance and individualitythis number of combi-
nations is far greater than the number of humans who have
ever lived on Earth! Genetic variation resulting from indepen-
dent assortment has been extremely important to the process
of evolution in all sexually reproducing organisms.
3.7 Laws of Probability Help
to Explain Genetic Events
Recall that genetic ratios are expressed as probabilitiesfor
example, 3/4 tall:1/4 dwarf. These values predict the outcome
of each fertilization event, such that the probability of each
zygote having the genetic potential for becoming tall is 3/4,
while the potential for becoming dwarf is 1/4. Probabilities
range from 0.0, when an event is certain not to occur, to 1.0,
when an event is certain to occur. When two or more events
occur independently but at the same time, we can calculate
the probability of possible outcomes when they occur togeth-
er. This is accomplished by applying the product lawthe
probability of two or more events occurring simultaneously is
equal to the product of their individual probabilities. Two or
more events are independent of one another if the outcome of
each one does not affect the outcome of any of the others
under consideration.
To illustrate the product law, consider the possible results if
you toss a penny (P) and a nickel (N) at the same time and
examine all combinations of heads (H) and tails (T) that can
occur. There are four possible outcomes:
(PH:NH) = (1>2)(1>2) = 1>4
(PT:NH) = (1>2)(1>2) = 1>4
(PH:NT) = (1>2)(1>2) = 1>4
(PT:NT) = (1>2)(1>2) = 1>4
The probability of obtaining a head or a tail in the toss of
either coin is 1/2 and is unrelated to the outcome of the toss of
Probability
Web Tutorial 3.3
the other coin. Thus, all four possible combinations are pre-
dicted to occur with equal probability.
If we want to calculate the probability where the possi-
ble outcomes of two events are independent of one
another but can be accomplished in more than one way,
we apply the sum law. For example, what is the proba-
bility of tossing our penny and nickel and obtaining one
head and one tail? In such a case, we do not care whether
it is the penny or the nickel that comes up heads, provid-
ed the other coin has the alternative outcome. As we saw
above, there are two ways in which the desired outcome
can be accomplished, each with a probability of 1/4.
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 50
50 Chapter 3 Mendelian Genetics
Thus, according to the sum law, the overall probability is
equal to
(1>4) + (1>4) = 1>2
One-half of all coin tosses are predicted to yield the desired
outcome.
These simple probability laws will be useful throughout our
discussions of transmission genetics and for solving genetics
problems. In fact, we already applied the product law when
we used the forked-line method to calculate the phenotypic
results of Mendels dihybrid and trihybrid crosses. When we
wish to know the results of a cross, we need only calculate the
probability of each possible outcome. The results of this cal-
culation then allow us to predict the proportion of offspring
expressing each phenotype or each genotype.
An important point to remember when you deal with proba-
bility is that predictions of possible outcomes are based on
large sample sizes. If we predict that 9/16 of the offspring of a
dihybrid cross will express both dominant traits, it is very
unlikely that, in a small sample, exactly 9 of every 16 off-
spring will express this phenotype. Instead, our prediction is
that, of a large number of offspring, approximately 9/16 of
them will do so. The deviation from the predicted ratio in
smaller sample sizes is attributed to chance, a subject we
examine in our discussion of statistics in the next section. As
you shall see, the impact of deviation due strictly to chance
diminishes as the sample size increases.
3.8 Chi-Square Analysis Evaluates
the Influence of Chance on
Genetic Data
Mendels 3:1 monohybrid and 9:3:3:1 dihybrid ratios are
hypothetical predictions based on the following assumptions:
(1) Each allele is dominant or recessive; (2) segregation is
operative; (3) independent assortment occurs; and (4) fertil-
ization is random. The final two assumptions are influenced
by chance events and are therefore subject to random fluctua-
tion. This concept of chance deviation is most easily illus-
trated by tossing a single coin numerous times and recording
the number of heads and tails observed. In each toss, there is a
probability of 1/2 that a head will occur and a probability of
1/2 that a tail will occur. Therefore, the expected ratio of
many tosses is 1:1. If a coin is tossed 1000 times, usually
about 500 heads and 500 tails will be observed. Any reason-
able fluctuation from this hypothetical ratio (e.g., 486 heads
and 514 tails) is attributed to chance.
As the total number of tosses is reduced, the impact of
chance deviation increases. For example, if a coin is tossed
only four times, you would not be too surprised if all four
tosses result in only heads or only tails. For 1000 tosses,
however, 1000 heads or 1000 tails would be most unexpect-
ed. In fact, you might believe that such a result would be
impossible. Actually, all heads or all tails in 1000 tosses can
be predicted to occur with a probability of (1/2)1000. Since
(1/2)20 is equivalent to less than one in a million times, an
event occurring with a probability of (1/2)1000 is virtually
impossible. Two major points are significant before we con-
sider chi-square analysis:
1. The outcomes of independent assortment and fertiliza-
tion, like coin tossing, are subject to random fluctuations
from their predicted occurrences as a result of chance
deviation.
2. As the sample size increases, the average deviation from
the expected results decreases. Therefore, a larger sample
size diminishes the impact of chance deviation on the
final outcome.
Chi-Square Calculations and the Null Hypothesis
In genetics, the ability to evaluate observed deviation is a
crucial skill. When we assume that data will fit a given ratio
such as 1:1, 3:1, or 9:3:3:1, we establish what is called the
null hypothesis 1H02. It is so named because the hypothe-
sis assumes that no real difference exists between measured
values (or ratio) and predicted values (or ratio). The appar-
ent difference can be attributed purely to chance. The null
hypothesis is evaluated using statistical analysis. On this
basis, the null hypothesis may either (1) be rejected or (2)
fail to be rejected. If it is rejected, the observed deviation
from the expected result is not attributed to chance alone.
The null hypothesis and the underlying assumptions lead-
ing to it must be reexamined. If the null hypothesis fails to
be rejected, any observed deviations are attributed to
chance.
One of the simplest statistical tests devised to assess the null
hypothesis is chi-square 1X 22 analysis. This test takes into
account the observed deviation in each component of an
expected ratio as well as the sample size and reduces them to
a single numerical value. The value for x2 is then used to esti-
mate how frequently the observed deviation can be expected
to occur strictly as a result of chance. The formula for chi-
square analysis is
1o - e22
x2 = g
e
where o is the observed value for a given category, e is the
expected value for that category, and g (the Greek letter
sigma) represents the sum of the calculated values for each
category of the ratio. Because (o - e) is the deviation (d) in
each case, the equation reduces to
d2
x2 = g
e
Table 3.1(a) shows a x2 calculation for the F2 results of a
hypothetical monohybrid cross. To analyze these data, you
work from left to right, calculating and entering the appropri-
ate numbers in each column. Regardless of whether the devi-
ation d is positive or negative, d2 always becomes positive
after the number is squared. In Table 3.1(b) the F2 results of a
hypothetical dihybrid cross are analyzed. Be sure that you
understand how each number was calculated in the dihybrid
example.
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 51

3.8 Chi-Square Analysis Evaluates the Influence of Chance on Genetic Data 51

TABLE 3.1 Chi-Square Analysis

(a) Monohybrid Cross

Expected Ratio Observed (o) Expected (e) Deviation (o  e) Deviation2 1d 22 d 2 /e

3>4 740 3>4(1000)
750 740  750
10 (10)2
100 100/750
0.13

1>4 260 1>4(1000)
250 260  250
10 (10)2
100 100/250
0.40

Total
1000 x2
0.53

p
0.48

(b) Dihybrid Cross

Expected Ratio (o) (e) (o  e) (d 2) d 2 /e

9>16 587 567 20 400 0.71

3>16 197 189 8 64 0.34

3>16 168 189 21 441 2.33

1>16 56 63 7 49 0.78

Total
1008 x2
4.16

p
0.26

The final step in chi-square analysis is to interpret the x2
value. To do so, you must initially determine the value of the
degrees of freedom (df), which is equal to n - 1, where n is
the number of different categories into which each datum point
may fall. For the 3:1 ratio, n = 2, so df = 2 - 1 = 1. For
the 9:3:3:1 ratio, n = 4 and df = 3. Degrees of freedom must
be taken into account because the greater the number of cate-
gories, the more deviation is expected as a result of chance.
Once you have determined the degrees of freedom, we can
interpret the x2 value in terms of a corresponding probability
value (p). Since this calculation is complex, we usually take
the p value from a standard table or graph. Figure 311
shows a wide range of x2 and p values for various degrees of
freedom in both a graph and a table. Lets use the graph to
determine the p value. The caption for Figure 311(b)
explains how to use the table.
To determine p, execute the following steps:
1. Locate the x2 value on the abscissa (the horizontal or
x-axis).
2. Draw a vertical line from this point up to the angled line
on the graph representing the appropriate df.
3. Extend a horizontal line from this point to the left until it
intersects the ordinate (the vertical or y-axis).
4. Estimate, by interpolation, the corresponding p value.
We used these steps for the monohybrid cross in Table 3.1(a)
to estimate the p value of 0.48 shown in Figure 311(a). For
the dihybrid cross, try this method to see if you can determine
the p value. Since the x2 value is 4.16 and df = 3, an
approximate p value is 0.26. Checking this result in the table
confirms that p values for both the monohybrid and dihybrid
crosses are between 0.20 and 0.50.
Interpreting Probability Values
So far, we have been concerned with calculating x2 values
and determining the corresponding p values. The most impor-
tant aspect of chi-square analysis is understanding the mean-
ing of the p value. Lets use the example of the dihybrid cross
in Table 3.1(b) (p = 0.26). In these discussions, it is simplest
to think of the p value as a percentage (e.g., 0.26 = 26%). In
our example, the p value indicates that if we repeat the same
experiment many times, 26 percent of the trials would be
expected to exhibit chance deviation as great or greater than
that seen in the initial trial. Conversely, 74 percent of the trials
would show less deviation than initially observed as a result
of chance. Thus, the p value reveals that a hypothesis (the
9:3:3:1 ratio in this case) is never proved or disproved
absolutely. Instead, a relative standard is set that enables us to
either reject or fail to reject the null hypothesisthis standard
is most often a p value of 0.05. When applied to chi-square
analysis, a p value less than 0.05 means that the observed
deviation in the set of results will be obtained by chance alone
less than 5 percent of the time. Such a p value indicates that
the difference between the observed and predicted results is
substantial and thus enables us to reject the null hypothesis.
On the other hand, p values of 0.05 or greater (0.05 to 1.0)
indicate that the observed deviation will be obtained by
chance alone 5 percent or more of the time. The conclusion is
not to reject the null hypothesis. Thus, for the p value of 0.26,
assessing the hypothesis that independent assortment
accounts for the results fails to be rejected. Therefore, the
observed deviation can be reasonably attributed to chance.
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 52

52 Chapter 3 Mendelian Genetics

(a) (b)
Degrees of freedom (df)
10 4 3 2 1 Probability (p)
0.0001
0.90 0.50 0.20 0.05 0.01 0.001

20 1 0.02 0.46 1.64 3.84 6.64 10.83

0.001 2 0.21 1.39 3.22 5.99 9.21 13.82
3 0.58 2.37 4.64 7.82 11.35 16.27
4 1.06 3.36 5.99 9.49 13.28 18.47
5 1.61 4.35 7.29 11.07 15.09 20.52
6 2.20 5.35 8.56 12.59 16.81 22.46
0.01 df 7 2.83 6.35 9.80 14.07 18.48 24.32
Probability (p)

30 8 3.49 7.34 11.03 15.51 20.09 26.13

0.03 9 4.17 8.34 12.24 16.92 21.67 27.88
0.05 10 4.87 9.34 13.44 18.31 23.21 29.59
15 8.55 14.34 19.31 25.00 30.58 37.30
0.1 25 16.47 24.34 30.68 37.65 44.31 52.62
50 37.69 49.34 58.16 67.51 76.15 86.60
0.2
x2 values

0.4
p = 0.48
Fails to reject the null hypothesis
0.7
Rejects the null hypothesis
1.0
40 30 20 15 10 7 5 43 2 1 0.1
x2 values

x2 = 0.53

FIGURE 3-11 (a) Graph for converting x2 values to p values. (b) Table of x2 values for selected values of df and p. x2 values that lead to
a p value of 0.05 or greater (darker blue areas) justify failure to reject the null hypothesis. Values leading to a p value of less than 0.05
(lighter blue areas) justify rejecting the null hypothesis. For example, using the table in part (b), where x2 = 0.53 for 1 degree of
freedom, the corresponding p value is between 0.20 and 0.50. The graph in (a) gives a more precise p value of 0.48 by interpolation.
Thus, we fail to reject the null hypothesis.

A final note is relevant here for the case where the null
hypothesis is rejected, that is, where p 0.05. Suppose we
are testing the null hypothesis that the data represented a
9:3:3:1 ratio, indicative of independent assortment. If the
null hypothesis is rejected, what are alternative interpreta-
tions of the data? Researchers will reassess the assumptions
that underlie the null hypothesis. In our example, we
assumed that segregation operates faithfully for both gene
pairs. We also assumed that fertilization is random and that
the viability of all gametes is equal regardless of geno-
typethat is, that all gametes are equally likely to partici-
pate in fertilization. Finally, following fertilization, we
assumed that all preadult stages and adult offspring are
equally viable regardless of their genotype. If any of these
assumptions is incorrect, the original hypothesis is not nec-
essarily invalid.
An example will clarify this. Suppose our null hypothesis is
that a dihybrid cross between fruit flies will result in 3/16
mutant wingless fly zygotes. However, not as many of the
mutant embryos may survive their preadult development or as
young adults, compared to flies whose genotype gives rise to
wings. As a result, when the data are gathered, there are fewer
than 3/16 wingless flies. Rejection of the null hypothesis
alone is not cause for us to disregard the validity of the postu-
lates of segregation and independent assortment, because
other factors are operative.
Now Solve This
Problem 18 on page 59 asks you to apply x2 analysis to
a set of data and determine whether the data fit sever-
al ratios.
Hint: In calculating x2, first determine the expected
outcomes using the predicted ratios. Then follow a
stepwise approach, determining the deviation in each
case, and calculating d 2>e for each category.
3.9 Pedigrees Reveal Patterns
of Inheritance of Human Traits
We now explore how to determine the mode of inheritance of
phenotypes in humans, where designed crosses are not possible
and where relatively few offspring are available for study. The
traditional way to study inheritance has been to construct a fami-
ly tree, indicating the presence or absence of the trait in question
for each member of each generation. Such a family tree is called
a pedigree. By analyzing a pedigree, we may be able to predict
how the trait under study is inheritedfor example, is it due to a
dominant or recessive allele? When many pedigrees for the same
trait are studied, we can often ascertain the mode of inheritance.
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 53

3.9 Pedigrees Reveal Patterns of Inheritance of Human Traits 53

horizontal line. Fraternal (or dizygotic) twins lack this con-

Female Male Sex unknown
Affected individuals
Parents (unrelated)
Consanguineous parents (related)
Offspring (in birth order)
1 2 3 4
Fraternal (dizygotic) twins
(sex may be the same or different)
Identical (monozygotic) twins
(sex must be the same)
4 4 Multiple individuals (unaffected)
P Proband (In this case, a male)
Deceased individual (In this case, a female)
Heterozygous carriers
I, II, III, etc. Successive generations
FIGURE 3-12 Conventions commonly encountered in human
pedigrees.
Pedigree Conventions
A variety of conventions are commonly used in pedigree con-
struction. Figure 312 illustrates a number of such conven-
tions: circles represent females and squares designate males.
Parents are connected by a single horizontal line, and vertical
lines lead to their offspring. If the parents are related
(consanguineous), such as first cousins, they are connected
by a double line. Offspring are called sibs (short for siblings)
and are connected by a horizontal sibship line. Sibs are
placed from left to right according to birth order, and are
labeled with Arabic numerals. Each generation is indicated by
a Roman numeral. If the sex of an individual is unknown, a
diamond is used. When a pedigree traces only a single trait,
the circles, squares, and diamonds are shaded if the phenotype
being considered is expressed and unshaded if not. In some
pedigrees, those individuals that fail to express a recessive
trait, but are known with certainty to be a heterozygous carri-
er, have a shaded dot within their unshaded circle or square. If
an individual is deceased and the phenotype is unknown, a
diagonal line is placed over the circle or square.
Twins are indicated by diagonal lines stemming from a
vertical line connected to the sibship line. For identical
(or monozygotic) twins, the diagonal lines are linked by a
necting line. A number within one of the symbols represents
numerous sibs of the same or unknown phenotypes. The indi-
vidual whose phenotype first brought attention to the investi-
gation and construction of the pedigree is called the proband
and is indicated by an arrow connected to the designation p.
This term applies to either a male or a female.
Pedigree Analysis
In Figure 313, two pedigrees are shown. The first illustrates a
representative pedigree for a trait that demonstrates autosomal
recessive inheritance, such as albinism. The male parent of the
first generation (I-1) is affected. Characteristic of a rare recessive
trait with an affected parent, the trait disappears in the offspring
of the next generation. Assuming recessiveness, we might predict
that the unaffected female parent (I-2) is a homozygous normal
individual because none of the offspring show the disorder. Had
she been heterozygous, one half of the offspring would be
expected to exhibit albinism, but none do. However, such a small
sample (three offspring) prevents us from knowing for certain.
Further evidence supports the prediction of a recessive trait. If
albinism were inherited as a dominant trait, individual II-3
would have to express the disorder in order to pass it to his off-
spring (III-3 and III-4), but he does not. Inspection of the off-
spring constituting the third generation (row III) provides still
further support for the hypothesis that albinism is a recessive
trait. If it is, parents II-3 and II-4 are both heterozygous, and
approximately one-fourth of their offspring should be affected.
Two of the six offspring do show albinism. This deviation from
the expected ratio is not unexpected in crosses with few off-
spring. Once we are confident that albinism is inherited as an
autosomal recessive trait, we could portray the II-3 and II-4 indi-
viduals with a shaded dot within their larger square and circle.
Finally, we can note that, characteristic of pedigrees for autoso-
mal traits, both males and females are affected with equal prob-
ability. In Chapter 4, we will examine a pedigree representing a
gene located on the sex-determining X chromosome. We will
see certain limitations imposed on the transmission of X-linked
traits, such as that these traits are more prevalent in male off-
spring and are never passed from affected fathers to their sons.
The second pedigree illustrates the pattern of inheritance for
a trait such as Huntington disease, which is caused by an auto-
somal dominant allele. The key to identifying such a pedigree
that reflects a dominant trait is that all affected offspring will
have a parent that also expresses the trait. It is also possible,
by chance, that none of the offspring will inherit the dominant
allele. If so, the trait will cease to exist in future generations.
Like recessive traits, provided that the gene is autosomal, both
males and females are equally affected.
When autosomal dominant diseases are rare within the
population, and most are, then it is highly unlikely that
affected individuals will inherit a copy of the mutant gene
from both parents. Therefore, in most cases, affected indi-
viduals are heterozygous for the dominant allele. As a result,
approximately one-half of the offspring inherit it. This is
borne out in the second pedigree in Figure 313. Further-
more, if a mutation is dominant, and a single copy is suffi-
cient to produce a mutant phenotype, homozygotes are
KLUGMC03_038-060HR 10/18/06 3:03 PM Page 54

54 Chapter 3 Mendelian Genetics

(a) Autosomal Recessive Trait FIGURE 3-13 Representative

pedigrees for two characteristics,
I Either I-3 or I-4 must be
each followed through three
heterozygous
1 2 3 4 generations.
Recessive traits typically
Analysis of Human Pedigrees

II skip generations
1 2 3 4 5 6 7
Web Tutorial 3.4

Recessive autosomal traits

III appear equally in both sexes
1 2 3 4 5 6
p

(b) Autosomal Dominant Trait

I I-1 is heterozygous for
a dominant allele
1 2
Dominant traits almost
II always appear in each
1 2 3 4 5 6 7 generation.
p
Affected individuals all have
III an affected parent.
Dominant autosomal traits
1 2 3 4 5 6 7 8 9 10 11
appear equally in both sexes.

likely to be even more severely affected, perhaps even fail- result. Such heterozygous individuals have heart attacks
ing to survive. An illustration of this is the dominant gene during the fourth decade of their life, or before. While het-
for familial hypercholesterolemia. Heterozygotes display a erozygotes have LDL levels about double that of a normal
defect in their receptors for low density lipoproteins, the so- individual, rare homozygotes have been detected. They lack
called LDLs. As a result, too little cholesterol is taken up by LDL receptors altogether, and have LDL levels nearly 10
cells from the blood, and elevated plasma levels of LDLs times above the normal range. They are likely to have a heart
attack very early in life, even before age five, and almost
inevitably before they reach the age of 20.
Pedigree analysis of many traits has historically been an
TABLE 3.2 Representative Recessive and extremely valuable research technique in human genetic
Dominant Human Traits studies. However, the approach does not usually provide the
certainty in drawing conclusions afforded by designed cross-
Recessive Traits Dominant Traits
es yielding large numbers of offspring. Nevertheless, when
Albinism Achondroplasia many independent pedigrees of the same trait or disorder are
analyzed, consistent conclusions can often be drawn. Table
Alkaptonuria Brachydactyly
3.2 lists numerous human traits and classifies them according
Ataxia telangiectasia Congenital stationary to their recessive or dominant expression. The genes control-
night blindness ling some of these traits are located on the sex-determining
Color blindness Ehler-Danlos syndrome chromosomes. We will discuss pedigrees for X-linked traits
in Chapter 4.
Cystic fibrosis Hypotrichosis

Duchenne muscular dystrophy Huntington disease

Galactosemia Hypercholesterolemia Now Solve This

Hemophilia Marfan syndrome Problem 26 on page 59 asks you to examine a pedigree
LeschNyhan syndrome Neurofibromatosis
for myopia and predict whether the trait is dominant or
recessive.
Phenylketonuria Phenylthiocarbamide
(PTC) tasting
Hint: One of the first things to look for are individuals
who express the trait, but neither of whose parents
Sicklecell anemia Porphyria also expresses the trait. Such an observation makes it
highly unlikely that the trait is dominant.
TaySachs disease Widows peak
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 55

Chapter Summary 55

GENETICS, Technology, and Society

Tay-Sachs Disease: The Molecular Basis of a Recessive Disorder in Humans

Tay-Sachs disease
(TSD) is an inherit-
ed disorder that causes unalterable destruc-
tion of the central nervous system. This
condition is particularly tragic because
infants with TSD are unaffected at birth and
appear to develop normally until they are
about six months old. Parents, believing
that they have a normal child, then must
witness the progressive loss of mental and
physical abilities. The disorder is severe, as
afflicted infants eventually become blind,
deaf, mentally retarded, and paralyzed,
often within only a year or two. Most do not
live beyond age five. Named for Warren
Tay and Bernard Sachs, who first described
the symptoms and associated them with the
disorder in the late 1800s, Tay-Sachs dis-
ease clearly demonstrates a pattern of auto-
somal recessive inheritance. Two
unaffected heterozygous parents, who most
often have no immediate family history of
the disorder, have a probability of one in
four of having a Tay-Sachs child. While
inheritance clearly displays a classic
Mendelian pattern, the protein product of
the affected gene has been identified, and
we now have a clear understanding of the
underlying molecular basis of the disorder.
TSD results from the loss of activity of a
single enzyme hexosaminidase A (Hex-
A). This enzyme is normally found in lyso-
somes, organelles that break down large
molecules for recycling by the cell. Hex-A
is needed to break down the ganglioside
GM2, a lipid component of nerve cell
membranes. Without functional Hex-A,
gangliosides accumulate within neurons in
the brain and cause deterioration of the ner-
vous system. Heterozygous carriers of TSD
with one normal copy of the gene produce
only about 50 percent of the normal amount
of Hex-A, but they show no symptoms of
the disorder. The observation that the activ-
ity of only one gene (one wild-type allele)
is sufficient for the normal development
and function of the nervous system explains
and illustrates the molecular basis of reces-
sive mutations. Only when both genes are
disrupted by mutation is the mutant pheno-
type evident.
The gene responsible for Tay-Sachs dis-
ease is located on chromosome 15 and
codes for the alpha subunit of the Hex-A
enzyme. Hex-A, like many enzymes, dis-
plays quaternary protein structure, consist-
ing of multiple polypeptide subunits. Since
the gene was isolated in 1985, more than
50 different mutations have been identified
that lead to TSD. Although the most com-
mon form of the disease is the infantile
form, where no functional Hex-A is pro-
duced, there is also a rare late-onset form
that occurs in patients with greatly reduced
Hex-A activity. Late-onset TSD is not
detectable until patients are in their twen-
ties or thirties, and it is generally much less
severe than the infantile form. Symptoms
include hand tremors, speech impediments,
muscle weakness, and loss of balance.
Tay-Sachs disease is almost a hundred
times more common in Ashkenazi Jews
Jews of central or eastern European
descentthan in the general population,
and also has a higher incidence in French
Canadians and in members of the Cajun
population in Louisiana. In the United
States, approximately one in every 27
Ashkenazi Jews is a heterozygous carrier
of TSD. By contrast, the carrier rate in the
general population and in Jews of
Sephardic (Spanish or Portuguese) origin
is approximately one in every 250.
Although there are currently no effective
treatments for TSD, recent advances in
carrier screening have helped to reduce the
prevalence of the disorder in high-risk
populations. Carriers can be identified by
tests that measure Hex-A activity or by
DNA-based tests that detect specific gene
mutations. When both parents are carriers,
prenatal diagnosis can be utilized with
each pregnancy in order to detect affected
fetuses. In addition, ganglioside synthesis
inhibitors and Hex-A enzyme replacement
therapy are currently being investigated as
potential treatments for TSD newborns.
Reference
Fernandes, F., and Shapiro, B. 2004.
Tay-Sachs Disease. Arch. Neurol.
61:146668.

CHAPTER SUMMARY
1. Over a century ago, Mendel studied inheritance patterns in the
garden pea and established the principles of transmission genetics.
2. Mendels postulates help describe the basis for the inheritance of
phenotypic expression. He showed that unit factors, later called
alleles, exist in pairs and exhibit a dominant/recessive relation-
ship in determining the expression of traits.
3. Mendel postulated that unit factors must segregate during ga-
mete formation, such that each gamete receives only one of the
two factors with equal probability.
4. Mendels postulate of independent assortment states that each
pair of unit factors segregates independently of other such pairs.
As a result, all possible combinations of gametes are formed
with equal probability.
5. The discovery of chromosomes in the late 1800s, along with
subsequent studies of their behavior during meiosis, led to the
rebirth of Mendels work, linking the behavior of his unit factors
to that of chromosomes during meiosis.
6. The Punnett square and the forked-line methods are used to pre-
dict the probabilities of phenotypes (and genotypes) from cross-
es involving two or more gene pairs.
7. Genetic ratios are expressed as probabilities. Thus, deriving out-
comes of genetic crosses requires an understanding of the laws
of probability.
8. Statistical analysis is used to test the validity of experimental
outcomes. In genetics, variations from the expected ratios due to
chance deviations can be anticipated.
9. Chi-square analysis allows us to assess the null hypothesis,
which states that there is no real difference between the expected
and observed values. As such, it tests the probability of whether
observed variations can be attributed to chance deviation.
10. Pedigree analysis is a method for studying the inheritance pat-
tern of human traits over several generations. It frequently pro-
vides the basis for determining the mode of inheritance of
human characteristics and disorders.
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 56

56 Chapter 3 Mendelian Genetics

KEY TERMS
albinism, 53 F2 generation (second filial), 40 proband, 53
allele, 41 gene, 41 product law, 44
branch diagram, 46 genotype, 41 Punnett square, 42
chance deviation, 50 heterozygote, 41 recessive, 41
chi-square 1x22 analysis, 50 hexosaminidase A (Hex-A), 55 reciprocal cross, 40
chromosomal theory of inheritance, 47 homozygote, 41 segregation, 41
consanguineous, 53 identical twins, 53 selfing, 40
continuous variation, 47 independent assortment, 45 sib (sibling), 53
degrees of freedom (df), 51 locus, 49 sibship line, 53
dihybrid cross, 43 maternal parent, 49 starch-branching enzyme (SBEI), 43
dinozygotic twins , 53 Mendels 9:3:3:1 dihybrid ratio, 46 sum law, 49
diploid number (2n), 48 monohybrid cross, 39 Tay-Sachs disease, 55
discontinuous variation, 47 monozygotic twins, 53 testcross, 43
dizygotic twins, 53 null hypothesis, 50 traits, 39
dominant, 41 paternal parent, 49 transmission genetics, 39
familial hypercholesterolemia, 54 pedigree, 52 transposable elements, 43
F1 generation (first filial), 40 phenotype, 41 trihybrid cross, 46
forked-line method, 46 P1 generation (parental), 40 unit factor, 41
fraternal twins, 53 probability value (p), 51 unit of inheritance, 39

Insights and Solutions

As a student, you will be asked to demonstrate your knowledge of Consider this problem: A recessive mutant allele, black, causes a
transmission genetics by solving genetics problems. Success at very dark body in Drosophila (a fruit fly) when homozygous. The
this task represents not only comprehension of theory but its wild-type (normal) color is gray. What F1 phenotypic ratio is pre-
application to more practical genetic situations. Most students dicted when a black female is crossed with a gray male whose
find problem solving in genetics to be challenging and rewarding. father was black?
This section will provide you with basic insights into the reason- To work out this problem, you must understand dominance and
ing essential to this process. recessiveness, as well as the principle of segregation. Further-
Genetics problems are in many ways similar to word problems more, you must use the information about the male parents
in algebra. The approach to solving them is identical: (1) Analyze father. Here is one way to solve this problem:
the problem carefully; (2) translate words into symbols, first 1. The female parent is black, so she must be homozygous for the
defining each one; and (3) choose and apply a specific technique mutant allele (bb).
to solve the problem. The first two steps are critical. The third
2. The male parent is gray; therefore, he must have at least one
step is largely mechanical.
dominant allele (B). His father was black (bb), and he received
The simplest problems state all necessary information about the one of the chromosomes bearing these alleles, so the male parent
P1 generation and ask you to find the expected ratios of the F1 and must be heterozygous (Bb).
F2 genotypes and/or phenotypes. Always follow these steps when
With this information, the problem is simple:
you encounter this type of problem:
1. Determine insofar as possible the genotypes of the individuals
in the P1 generation. bb
Bb
2. Determine what gametes may be formed by the P1 parents. Homozygous Heterozygous
3. Recombine gametes by the Punnett square or the forked-line black female gray male
methods, or if the situation is very simple, by inspection. Read
the F1 phenotypes.
B b 1/2 Heterozygous gray
4. Repeat the process to obtain information about the F2 males and females, Bb
generation. b Bb bb F1
1/2 Homozygous black
Determining the genotypes from the given information requires males and females, bb
that you understand the basic theory of transmission genetics.
KLUGMC03_038-060HR 10/18/06 3:03 PM Page 57

Insights and Solutions 57

Apply this approach to the following problems. F1: 3>8 full, round
1. Mendel found that full pods are dominant over constricted 3>8 full, wrinkled
pods while round seeds are dominant over wrinkled seeds. One 1>8 constricted, round
of his crosses was between full, round plants and constricted,
1>8 constricted, wrinkled
wrinkled plants. From this cross, he obtained an F1 generation
that was all full and round. In the F2 generation, Mendel Determine the genotypes and phenotypes of the parents.
obtained his classic 9:3:3:1 ratio. Using this information, Solution: This problem is more difficult and requires keener insight
determine the expected F1 and F2 results of a cross between because you must work backward. The best approach is to consider
homozygous constricted, round plants and full, wrinkled the outcomes of pod shape separately from those of seed texture.
plants. Of all the plants, 3/8 + 3/8 = 3/4 are full and 1/8 + 1/8 = 1/4
Solution: Define gene symbols for each pair of contrasting traits. are constricted. Of the various genotypic combinations that can serve
Use the lowercase first letter of the recessive traits to designate as parents, which combination will give rise to a ratio of 3/4:1/4?
those phenotypes and the uppercase first letter to designate the This ratio is identical to Mendels monohybrid F2 results, and we can
dominant traits. Thus, C and c indicate full and constricted, and propose that both unknown parents share the same genetic character-
W and w indicate round and wrinkled phenotypes, respectively. istic as the monohybrid F1 parents; they must both be heterozygous
Determine the genotypes of the P 1 generation, form the for the genes controlling pod shape and thus are Cc.
gametes, reconstitute the F 1 generation, and read off the Before we accept this hypothesis, lets consider the possible genotyp-
phenotype(s): ic combinations that control seed texture. If we consider this character-
istic alone, we see that the traits are expressed in a ratio of
P1: ccWW CCww 3/8 + 1/8 = 1/2 round: 3/8 + 1/8 = 1/2 wrinkled. To generate
*
constricted, round full, wrinkled such a ratio, the parents cannot both be heterozygous, or their offspring
p p would yield a 3/4 : 1/4 phenotypic ratio. They cannot both be
Gametes: cW Cw homozygous, or all of their offspring would express a single phenotype.
F1: CcWw Thus, we are left with testing the hypothesis that one parent is homozy-
full, round
gous and one is heterozygous for the alleles controlling texture. The
potential case of WW * Ww does not work, since it yields only a sin-
You can immediately see that the F1 generation expresses
gle phenotype. This leaves us with the potential case of Ww * ww.
both dominant phenotypes and is heterozygous for both gene
Offspring in such a mating will yield 1/2 Ww (round):1/2 ww (wrin-
pairs. Thus, you expect that the F 2 generation will yield the
kled), exactly the outcome we are seeking.
classic Mendelian ratio of 9:3:3:1. Lets work it out anyway
just to confirm this, using the forked-line method. Both gene Now, lets combine the hypotheses and predict the outcome of the
pairs are heterozygous and can be expected to assort inde- cross. In our solution, we use a dash () to indicate that the sec-
pendently, so we can predict the F 2 outcomes from each gene ond allele may be either dominant or recessive, since we are only
pair separately and then proceed with the forked-line predicting phenotypes.
method. 1>2 Ww : 3>8 CWw full, round
Every F2 offspring is subject to the following probabilities: 3>4 C
1>2 ww : 3>8 Cww full, wrinkled
Cc * Cc Ww * Ww 1>2 Ww : 1>8 ccWw constricted, round
1>4 cc
p p 1>2 ww : 1>8 ccww constricted, wrinkled
CC WW As you can see, this cross produces offspring according to our
Cc s full Ww s round initial information, and we have solved the problem. Note that in
this solution, we used genotypes in the forked-line method, in
cC wW
contrast to the use of phenotypes in the earlier solution.
cc constricted ww wrinkled 3. Determine the probability that a plant of genotype CcWw will be
produced from parental plants with the genotypes CcWw and Ccww.
The forked-line method then confirms the 9:3:3:1 phenotypic
ratio. Remember that this represents proportions of 9/16:3/16: Solution: The two gene pairs demonstrate straightforward domi-
3/16:1/16. Note that we are applying the product law as we com- nance and recessiveness and assort independently during gamete
pute the final probabilities: formation. We need only calculate the individual probabilities of AU: Please
obtaining the two separate outcomes (Cc and Ww) and apply the confirm that
13>4213>42 product law to calculate the final probability:
3>4 round " 9>16 full, round these divisions
3>4 full Cc * Cc 1>4 CC : 1/2 Cc : 1>4 cc are ratio. If so,
13>4211>42 " 3>16 full, wrinkled Ww * ww 1>2 Ww : 1>2 ww there shouldn't
1>4 wrinkled be any space
p = (1>2 Cc)(1>2 Ww) = 1>4 CcWw
11>4213>42 before and after
3>4 round " 3>16 constricted, round 4. In the laboratory, a genetics student crossed flies that had
colon.
normal, long wings with flies expressing the dumpy mutation
1>4 constricted
11>4211>42 " 1>16 constricted, wrinkled
(truncated wings), which she believed was a recessive trait. In
1>4 wrinkled the F1 generation, all flies had long wings. The following
results were obtained in the F2 generation:
2. In another cross involving parent plants of unknown genotype 792 long-winged flies
and phenotype, the following offspring were obtained. 208 dumpy-winged flies
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 58

58 Chapter 3 Mendelian Genetics

The student tested the hypothesis that the dumpy wing is inherit- We consult Figure 311 to determine the probability (p) and
ed as a recessive trait, using chi-square analysis of the F2 data. determine whether the deviations can be attributed to chance.
(a) What ratio was hypothesized? There are two possible outcomes (n), so the degrees of freedom
(df) = n - 1 or 1. The table in Figure 311(b) shows that p is a
(b) Did the analysis support the hypothesis?
value between 0.01 and 0.001; the graph in Figure 311(a) gives
(c) What do the data suggest about the dumpy mutation? an estimate of about 0.001. Since p 6 0.05, we reject the null
Solution: hypothesis. The data do not fit a 3:1 ratio.
(a) The student hypothesized that the F2 data (792:208) fit (c) When we accepted Mendels 3:1 ratio as a valid expression of
Mendels 3:1 monohybrid ratio for recessive genes. the monohybrid cross, numerous assumptions were made. Exam-
(b) The initial step in x2 analysis is to calculate the expected ining our underlying assumptions may explain why the null
results (e) if the ratio is 3:1. Then we can compute deviation hypothesis was rejected. We assumed that all genotypes are
o - e (d) and the remaining numbers. equally viablethat genotypes yielding long wings are equally
likely to survive from fertilization through adulthood as the geno-
type yielding dumpy wings. Further study may reveal that
Ratio o e d d2 d2 /e dumpy-winged flies are somewhat less viable than normal flies.
As a result, we would expect less than 1>4 of the total offspring to
3>4 792 750 42 1764 2.35
express dumpy wings. This observation is borne out in the data,
1>4 208 250 -42 1764 7.06 although we have not proven that this is true.
Total = 1000
d2
x2 = a
e
= 2 .35 + 7.06
= 9 .41

PROBLEMS AND DISCUSSION QUESTIONS

When working out genetics problems in this and succeeding Then F1 offspring were selectively mated with the following re-
chapters, always assume that members of the P1 generation are sults. (The P1 cross giving rise to each F1 pigeon is indicated in
homozygous, unless the information given, or the data, indicates parentheses.)
otherwise.
F2 Progeny
1. In a cross between a black and a white guinea pig, all members
of the F1 generation are black. The F2 generation is made up of F1 F1 Crosses Checkered Plain
approximately 3>4 black and 1>4 white guinea pigs. Diagram
(d) checkered (a) * plain (c) 34 0
this cross, and show the genotypes and phenotypes.
2. Albinism in humans is inherited as a simple recessive (e) checkered (b) * plain (c) 17 14
trait. Determine the genotypes of the parents and off- (f) checkered (b) * checkered (b) 28 9
spring for the following families. When two alternative (g) checkered (a) * checkered (b) 39 0
genotypes are possible, list both. (a) Two nonalbino (nor-
mal) parents have five children, four normal and one albi-
no. (b) A normal male and an albino female have six How are the checkered and plain patterns inherited? Predict the
children, all normal. results of the F1 * F1 mating from cross (b).
3. In a problem involving albinism (see Problem 2), which of
Mendels postulates are demonstrated?
4. Why was the garden pea a good choice as an experimental or-
ganism in Mendels work?
5. Pigeons exhibit a checkered or plain feather pattern. In a series
of controlled matings, the following data were obtained:

F1 Progeny
P1 Cross Checkered Plain 6. Mendel crossed peas having round seeds and yellow cotyledons
(a) checkered * checkered 36 0 with peas having wrinkled seeds and green cotyledons. All the
F1 plants had round seeds with yellow cotyledons. Diagram this
(b) checkered * plain 38 0
cross through the F2 generation, using both the Punnett square
(c) plain * plain 0 35 and forked-line methods.
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 59

Problems and Discussion Questions 59

7. Determine the genotypes of the parental plants by analyzing the 18. In one of Mendels dihybrid crosses, he observed 315 round, yel-
phenotypes of the offspring from these crosses: low; 108 round, green; 101 wrinkled, yellow; and 32 wrinkled,
green F2 plants. Analyze these data using chi-square analysis to
Parental Plants Offspring see whether (a) they fit a 9:3:3:1 ratio; (b) the round, wrinkled
traits fit a 3:1 ratio; or (c) the yellow, green traits fit a 3:1 ratio.
(a) round, yellow * round, yellow 3/4 round, yellow 19. A geneticist, in assessing data that fell into two phenotypic class-
1/4 wrinkled, yellow es, observed values of 250:150. He decided to perform chi-square
(b) round, yellow * wrinkled, yellow 6/16 wrinkled, yellow analysis using two different null hypotheses: (a) The data fit a 3:1
ratio; and (b) the data fit a 1:1 ratio. Calculate the x2 values for
2/16 wrinkled, green
each hypothesis. What can you conclude about each hypothesis?
6/16 round, yellow 20. The basis for rejecting any null hypothesis is arbitrary. The
2/16 round, green researcher can set more or less stringent standards by deciding to
(c) round, yellow * wrinkled, green 1/4 round, yellow raise or lower the critical p value. Would the use of a standard of
p = 0.10 be more or less stringent in failing to reject the null
1/4 round, green
hypothesis? Explain.
1/4 wrinkled, yellow 21. Consider three independently assorting gene pairs, A/a, B/b,
1/4 wrinkled, green and C/c, where each demonstrates typical dominance (A, B,
C), and recessiveness (aa, bb, cc). What is the probability of
8. Are any of the crosses in Problem 7 testcrosses? If so, which obtaining an offspring that is AABbCc from parents that are
one(s)? AaBbCC and AABbCc?
9. Which of Mendels postulates can be demonstrated in the cross- 22. What is the probability of obtaining a triply recessive individual
es of Problem 7 but not in those in Problems 1 and 5? State this from the parents shown in Problem 21?
postulate. 23. Of all offspring of the parents in Problem 21, what proportion
10. Correlate Mendels four postulates with what is now known will express all three dominant traits?
about homologous chromosomes, genes, alleles, and the process 24. For the following pedigree, predict the mode of inheritance and
of meiosis. the resulting genotypes of each individual. Assume that the alle-
11. What is the basis for homology among chromosomes? les A and a control the expression of the trait.
12. Distinguish between homozygosity and heterozygosity.
13. In Drosophila, gray body color is dominant over ebony body l
color, while long wings are dominant over vestigial wings. Work 1 2 3 4
the following crosses through the F2 generation, and determine
the genotypic and phenotypic ratios for each generation. Assume ll
that the P1 individuals are homozygous: 1 2 3 4 5 6 7 8
(a) gray, long * ebony, vestigial
(b) gray, vestigial * ebony, long lll
(c) gray, long * gray, vestigial 1 2 3 4 5 6
14. How many different types of gametes can be formed by individ-
uals of the following genotypes? What are they in each case? (a) lV
AaBb, (b) AaBB, (c) AaBbCc, (d) AaBBcc, (e) AaBbcc, and (f) 1 2 3 4 5 6 7
AaBbCcDdEe?
15. Using the forked-line method, determine the genotypic and phe-
notypic ratios of these trihybrid crosses: 25. Which of Mendels postulates are demonstrated by the pedigree
(a) AaBbCc * AaBBCC, (b) AaBBCc * aaBBCc, and (c) in Problem 24? List and define these postulates.
AaBbCc * AaBbCc. 26. The following pedigree follows the inheritance of myopia (near-
16. Mendel crossed peas with round, green seeds with peas hav- sightedness) in humans. Predict whether the disorder is inherited
ing wrinkled, yellow seeds. All F1 plants had seeds that were as a dominant or a recessive trait. Based on your prediction, in-
round and yellow. Predict the results of testcrossing these F1 dicate the most probable genotype for each individual.
plants.
17. Shown are F2 results of two of Mendels monohybrid crosses.
State a null hypothesis that you will test using chi-square analy-
sis. Calculate the x2 value and determine the p value for both
crosses, then interpret the p values. Which cross shows a greater
amount of deviation?
(a) Full pods 882
Constricted pods 299
(b) Violet flowers 705 27. Draw all possible conclusions concerning the mode of inheri-
White flowers 224 tance of the trait expressed in each of the following limited pedi-
grees. (Each case is based on a different trait.)
KLUGMC03_038-060HR 10/18/06 1:01 PM Page 60

60 Chapter 3 Mendelian Genetics

(a) (a) Using standard pedigree symbols, draw a pedigree of these

individuals families, showing the relevant individuals.
(b) The couple asks you to calculate the probability that they
both are heterozygous for the TSD allele.
(c) They also want to know the probability that neither of them is
heterozygous.
(b) (d) They also ask you for the probability that one of them is het-
erozygous but the other is not.
[Hint: The answers to (b), (c), and (d) should add up to 1.0.]
30. The wild-type (normal) fruit fly, Drosophila melanogaster, has
straight wings and long bristles. Mutant strains have been isolat-
ed with either curled wings or short bristles. The genes repre-
(c) senting these two mutant traits are located on separate
chromosomes. Carefully examine the data from the five crosses
below. (a) For each mutation, determine whether it is dominant
or recessive. In each case, identify which crosses support your
answer; and (b) define gene symbols, and determine the geno-
(d) types of the parents for each cross.

Number of Progeny
straight straight curled curled
wings, wings, wings, wings,
long short long short
Cross bristles bristles bristles bristles
1 straight, short *
28. Two true-breeding pea plants are crossed. One parent is round, straight, short 30 90 10 30
terminal, violet, constricted, while the other expresses the con-
2 straight, long *
trasting phenotypes of wrinkled, axial, white, full. The four pairs
straight, long 120 0 40 0
of contrasting traits are controlled by four genes, each located on
a separate chromosome. In the F1 generation, only round, axial, 3 curled, long *
violet, and full are expressed. In the F2 generation, all possible straight, short 40 40 40 40
combinations of these traits are expressed in ratios consistent 4 straight, short *
with Mendelian inheritance. straight, short 40 120 0 0
(a) What conclusion can you draw about the inheritance of these 5 curled, short *
traits based on the F1 results? straight, short 20 60 20 60
(b) Which phenotype appears most frequently in the F2 results?
Write a mathematical expression that predicts the frequency 31. To assess Mendels law of segregation using tomatoes, a true-
of occurrence of this phenotype. breeding tall variety (SS) is crossed with a true-breeding short
(c) Which F2 phenotype is expected to occur least frequently? variety (ss). The heterozygous tall plants (Ss) were crossed to
Write a mathematical expression that predicts this frequency. produce the two sets of F2 data shown below:
(d) How often is either P1 phenotype likely to occur in the F2
generation?
Set I Set II
(e) If the F1 plant is testcrossed, how many different phenotypes
will be produced, and how does this number compare to the 30 tall 300 tall
number of different phenotypes in the F2 generation dis- 5 short 50 short
cussed in part (b)?
29. Tay-Sachs disease (TSD) is an inborn error of metabolism that (a) Using chi-square analysis, analyze the results for both data
results in death, usually before the age of five. You are a genetic sets. Calculate x2 values, and estimate the p values in both
counselor, and you interview a phenotypically normal couple cases.
who consult you because the man had a female first cousin (on (c) From the analysis in part (a), what can you conclude about
his fathers side) who died from TSD, and the woman had a the importance of generating large data sets in experimental
maternal uncle with TSD. There are no other known cases in ei- settings?
ther family, and none of the matings were/are between related
individuals. Assume that this trait is rare in this population.
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 136

C H A P T E R

7
Linkage and
Chromosome Mapping
in Eukaryotes

Chiasmata between synapsed

homologs during the first
meiotic prophase of meiosis.

CHAPTER CONCEPTS
Chromosomes in eukaryotes contain many
genes whose locations are fixed along the
length of the chromosomes.
Unless separated by crossing over,
alleles present on a chromosome
segregate as a unit during gamete
formation.
Crossing over between homologs during Genetic maps depict relative locations of
meiosis creates recombinant gametes with genes on chromosomes in a species.
different combinations of alleles that
enhance genetic variation.
Crossing over between homologs serves as
the basis for the construction of
chromosome maps.

136
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 137

7.1 Genes Linked on the Same Chromosome Segregate Together 137

W
alter Sutton, along with Theodor Boveri, was instru-
mental in uniting the fields of cytology and genet- 7.1 Genes Linked on the Same

Linkage and Recombination

Web Tutorial 7.1
ics. As early as 1903, Sutton pointed out the Chromosome Segregate Together
likelihood that there must be many more unit factors than
A simplified overview of the major theme of this chapter is
chromosomes in most organisms. Soon thereafter, genetics
given in Figure 71, which contrasts the meiotic conse-
investigations revealed that certain genes segregate as if they
quences of (a) independent assortment, (b) linkage without
were somehow joined or linked together. Further investiga-
crossing over, and (c) linkage with crossing over. In Figure
tions showed that such genes are part of the same chromo-
71(a), we see the results of independent assortment of two
some and may indeed be transmitted as a single unit. We now
pairs of chromosomes, each containing one heterozygous
know that most chromosomes contain a very large number of
gene pair. No linkage is exhibited. When a large number of
genes. Those that are part of the same chromosome are said to
meiotic events are observed, four genetically different
be linked and to demonstrate linkage in genetic crosses.
gametes are formed in equal proportions, and each contains a
Because the chromosome, not the gene, is the unit of trans-
different combination of alleles of the two genes.
mission during meiosis, linked genes are not free to undergo
Now lets compare these results with what occurs if the same
independent assortment. Instead, the alleles at all loci of one
genes are linked on the same chromosome. If no crossing over
chromosome should, in theory, be transmitted as a unit during
occurs between the two genes [Figure 71(b)] only two genetical-
gamete formation. However, in many instances this does
ly different gametes are formed. Each gamete receives the alleles
not occur. During the first meiotic prophase, when homologs
present on one homolog or the other, which is transmitted intact as
are paired or synapsed, a reciprocal exchange of chromosome
the result of segregation. This case demonstrates complete link-
segments can take place. This crossing over results in the
age, which produces only parental or noncrossover gametes.
reshuffling, or recombination, of the alleles between
The two parental gametes are formed in equal proportions.
homologs and always occurs during the tetrad stage.
Though complete linkage between two genes seldom occurs, it is
The degree of crossing over between any two loci on a sin-
useful to consider the theoretical consequences of this concept.
gle chromosome is proportional to the distance between them.
Figure 71(c) shows the results of crossing over between two
Therefore, depending on which loci are being studied, the
linked genes. As you can see, this crossover involves only two
percentage of recombinant gametes varies. This correlation
nonsister chromatids of the four chromatids present in the
allows us to construct chromosome maps, which give the rel-
tetrad. This exchange generates two new allele combinations,
ative locations of genes on chromosomes.
called recombinant or crossover gametes. The two chro-
In this chapter, we will discuss linkage, crossing over, and
matids not involved in the exchange result in noncrossover
chromosome mapping in more detail. We will conclude by
gametes, like those in Figure 71(b). The frequency with which
entertaining the rather intriguing question of why Mendel,
crossing over occurs between any two linked genes is generally
who studied seven genes in an organism with seven chromo-
proportional to the distance separating the respective loci along
somes, did not encounter linkage. Or did he?
the chromosome. In theory, two randomly selected genes can
be so close to each other that crossover events are too infre-
quent to easily detect. As shown in Figure 71(b), this complete
How Do We Know? linkage produces only parental gametes. On the other hand, if a
small but distinct distance separates two genes, few recombi-
In this chapter, we will focus on genes located on the same nant and many parental gametes will be formed. As the distance
eukaryotic chromosome, a concept called linkage. We are par- between the two genes increases, the proportion of recombinant
ticularly interested in how such linked genes are transmitted gametes increases and that of the parental gametes decreases.
to offspring and how the results of such transmission enable As we will discuss later in this chapter, when the loci of two
us to map them. As you study this topic, you should try to an- linked genes are far apart, the number of recombinant
swer several fundamental questions: gametes approaches, but does not exceed, 50 percent. If
50 percent recombinants occur, the result is a 1:1:1:1 ratio of
1. How do we know that specific genes are linked on a single
chromosome, in contrast to being located on separate
chromosomes?
2. How was it determined that linked genes on homologous
chromosomes are recombined during meiosis? Now Solve This
3. How do we know the order of, and intergenic distance be- Problem 9 on page 160 asks you to contrast the results
tween, genes found on the same chromosome? of a testcross when two genes are unlinked versus
linked, and when they are linked, if they are very far
4. In organisms where designed matings do not occur (such apart or relatively close together.
as humans), how do we know that genes are linked, and
how do we map them? Hint: The results are indistinguishable when two
genes are unlinked compared to the case where they
5. How do we know that crossing over results from a physical are linked but so far apart that crossing over always
exchange between chromatids? intervenes between them during meiosis.
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 138

138 Chapter 7 Linkage and Chromosome Mapping in Eukaryotes

(a) Independent assortment: Two genes on
two different homologous pairs of chromosomes
A B
A B
a b
a b
the four types (two parental and two recombinant gametes). In
this case, transmission of two linked genes is indistinguish-
able from that of two unlinked, independently assorting
genes. That is, the proportion of the four possible genotypes is
identical, as shown in Figures 71(a) and (c).

The Linkage Ratio

If complete linkage exists between two genes because of their
close proximity, and organisms heterozygous at both loci are
A A
mated, a unique F2 phenotypic ratio results, which we desig-
B b nate the linkage ratio. To illustrate this ratio, lets consider a
cross involving the closely linked, recessive, mutant genes
Gametes brown eye (bw) and heavy wing vein (hv) in Drosophila
a a melanogaster (Figure 72). The normal, wild-type alleles
bw + and hv + are both dominant and result in red eyes and thin
B b
wing veins, respectively.
In this cross, flies with mutant brown eyes and normal thin
wing veins are mated to flies with normal red eyes and mutant
(b) Linkage: Two genes on a single pair heavy wing veins. In more concise terms, brown-eyed flies
of homologs; no exchange occurs
are crossed with heavy-veined flies. If we extend the system
A B of genetic symbols established in Chapter 4, linked genes are
A B
a b represented by placing their allele designations above and
a b below a single or double horizontal line. Those above the line
are located at loci on one homolog, and those below are locat-
ed at the homologous loci on the other homolog. Thus, we
represent the P1 generation as follows:
A B A B bw hv + bw + hv
P1: *
bw hv + bw + hv

Gametes brown, thin red, heavy

These genes are located on an autosome, so no distinction
a b a b
between males and females is necessary.
In the F1 generation, each fly receives one chromosome of
each pair from each parent. All flies are heterozygous for both
gene pairs and exhibit the dominant traits of red eyes and thin
(c) Linkage: Two genes on a single pair wing veins:
of homologs; exchange occurs between
two nonsister chromatids bw hv +
F1:
A B bw + hv
A B Nonsister
a b chromatids red, thin
a b
As shown in Figure 72(a), when the F1 generation is inter-
bred, each F1 individual forms only parental gametes because
of complete linkage. After fertilization, the F2 generation is
A B A b produced in a 1:2:1 phenotypic and genotypic ratio. One-
fourth of this generation shows brown eyes and thin wing
Noncrossover Crossover veins; one-half shows both wild-type traits, namely, red eyes
gamete gamete
and thin wing veins; and one-fourth shows red eyes and
Gametes heavy wing veins. In more concise terms, the ratio is 1 brown:
a B a b 2 wild:1 heavy. Such a 1:2:1 ratio is characteristic of com-
plete linkage. Complete linkage is usually observed only
Crossover Noncrossover
gamete gamete when genes are very close together and the number of proge-
ny is relatively small.
FIGURE 71 Results of gamete formation where two heterozygous Figure 72(b) also gives the results of a testcross with the F1
genes are (a) on two different pairs of chromosomes; (b) on the flies. Such a cross produces a 1:1 ratio of brown, thin and red,
same pair of homologs, but where no exchange occurs between heavy flies. Had the genes controlling these traits been incom-
them; and (c) on the same pair of homologs, where an exchange pletely linked or located on separate autosomes, the testcross
occurs between two nonsister chromatids. would have produced four phenotypes rather than two.
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 139

7.1 Genes Linked on the Same Chromosome Segregate Together 139

hv bw hv bw
P1

hv bw hv bw
brown eyes, red eyes,
thin veins heavy veins

Gamete formation

hv bw hv bw

Testcross parent

hv bw hv bw
F1
Because of
complete linkage, hv bw hv bw
F1 individuals red eyes, brown eyes,
form only thin veins heavy veins
parental gametes.

Gamete formation

hv bw hv bw hv bw

(a) F1
F1 (b) F1
Testcross parent

hv bw hv bw hv bw

hv bw hv bw hv bw

hv bw hv bw

hv bw hv bw hv bw
brown, thin red, thin brown, thin
hv bw hv bw hv bw
hv bw hv bw

hv bw hv bw hv bw
red, thin red, heavy red, heavy

F2 progeny Testcross progeny

1/4 brown, thin:2/4 red, thin:1/4 red, heavy 1/2 brown, thin:1/2 red, heavy
1:2:1 ratio 1:1 ratio
FIGURE 72 Results of a cross involving two genes located on the same chromosome where complete linkage is demonstrated.
(a) The F2 results of the cross. (b) The results of a testcross involving the F1 progeny.
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 140
140 Chapter 7 Linkage and Chromosome Mapping in Eukaryotes
When large numbers of mutant genes present in any given
species are investigated, genes located on the same chromo-
some show evidence of linkage to one another. As a result,
linkage groups can be established, one for each chromo-
some. In theory, the number of linkage groups should corre-
spond to the haploid number of chromosomes. In diploid
organisms in which large numbers of mutant genes are avail-
able for genetic study, this correlation has been confirmed.
Genes and Mapping
Web Tutorial 7.2
7.2 Crossing Over Serves as the Basis
of Determining the Distance
Between Genes During Mapping
It is highly improbable that two randomly selected genes
linked on the same chromosome will be so close to one
another along the chromosome that they demonstrate com-
plete linkage. Instead, crosses involving two such genes
almost always produce a percentage of offspring resulting
from recombinant gametes. This percentage is variable and
depends on the distance between the two genes along the
chromosome. This phenomenon was first explained around
1910 by two Drosophila geneticists, Thomas H. Morgan and
his undergraduate student, Alfred H. Sturtevant.
Morgan, Sturtevant, and Crossing Over
In his studies, Morgan investigated numerous Drosophila
mutations located on the X chromosome. When he analyzed
crosses involving only one trait, he deduced the mode of
X-linked inheritance. However, when he made crosses involv-
ing two X-linked genes, his results were initially puzzling.
For example, female flies expressing the mutant yellow body
(y) and white eyes (w) alleles were crossed with wild-type
males (gray bodies and red eyes). The F1 females were wild
type, while the F1 males expressed both mutant traits. In the
F2 , the vast majority of the offspring showed the expected
parental phenotypeseither yellow-bodied, white-eyed flies
or wild-type flies (gray-bodied, red-eyed). However, the
remaining flies, less than 1.0 percent, were either yellow-
bodied with red eyes or gray-bodied with white eyes. It was as
if the two mutant alleles had somehow separated from each
other on the homolog during gamete formation in the F1
female flies. This cross is illustrated in cross A of Figure 73,
using data later compiled by Sturtevant.
When Morgan studied other X-linked genes, the same basic
pattern was observed, but the proportion of the unexpected F2
phenotypes differed. For example, in a cross involving the
mutant white eye, miniature wing (m) alleles, the majority of
the F2 again showed the parental phenotypes, but a much
higher proportion of the offspring appeared as if the mutant
genes had separated during gamete formation. This is illus-
trated in cross B of Figure 73, again using data subsequently
compiled by Sturtevant.
Morgan was faced with two questions: (1) What was the
source of gene separation, and (2) why did the frequency
of the apparent separation vary depending on the genes
being studied? The answer he proposed for the first ques-
tion was based on his knowledge of earlier cytological
observations made by F. A. Janssens and others. Janssens
had observed that synapsed homologous chromosomes in
meiosis wrapped around each other, creating chiasmata
(sing., chiasma) where points of overlap are evident (see
the photo on p. 136). Morgan proposed that chiasmata
could represent points of genetic exchange.
In the crosses shown in Figure 73, Morgan postulated that
if an exchange occurs during gamete formation between the
mutant genes on the two X chromosomes of the F1 females,
the unique phenotypes will occur. He suggested that such
exchanges led to recombinant gametes in both the
yellowwhite cross and the whiteminiature cross, in contrast
to the parental gametes that have undergone no exchange.
On the basis of this and other experiments, Morgan concluded
that linked genes exist in a linear order along the chromosome
and that a variable amount of exchange occurs between any
two genes during gamete formation.
In answer to the second question, Morgan proposed that two
genes located relatively close to each other along a chromo-
some are less likely to have a chiasma form between them
than if the two genes are farther apart on the chromosome.
Therefore, the closer two genes are, the less likely a genetic
exchange will occur between them. Morgan was the first to
propose the term crossing over to describe the physical
exchange leading to recombination.
Sturtevant and Mapping
Morgans student, Alfred H. Sturtevant, was the first to realize
that his mentors proposal could be used to map the sequence
of linked genes. According to Sturtevant,
In a conversation with Morgan I suddenly realized
that the variations in strength of linkage, already attrib-
uted by Morgan to differences in the spatial separation
of the genes, offered the possibility of determining
sequences in the linear dimension of a chromosome. I
went home and spent most of the night (to the neglect of
my undergraduate homework) in producing the first
chromosomal map.
Sturtevant compiled data from numerous crosses made by
Morgan and other geneticists involving recombination
between the genes represented by the yellow, white, and
miniature mutants. These data are shown in Figure 73. The
following recombination between each pair of these three
genes, published in Sturtevants paper in 1913, is as follows:
(1) yellowwhite 0.5%
(2) whiteminiature 34.5%
(3) yellowminiature 35.4%
Because the sum of (1) and (2) approximately equals (3),
Sturtevant suggested that the recombination frequencies
between linked genes are additive. On this basis, he predicted
that the order of the genes on the X chromosome is
yellowwhiteminiature. In arriving at this conclusion, he rea-
soned as follows: The yellow and white genes are apparently
close to each other because the recombination frequency is
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 141

7.2 Crossing Over Serves as the Basis of Determining the Distance Between Genes During Mapping 141

Cross A Cross B
y w y w w m w m

P1

y w w m
yellow, white wild type white, miniature wild type

y w y w w m w m



F1
y w w m
wild type yellow, white wild type white, miniature

Parental Recombinant Parental Recombinant

types (99.5%) types (0.5%) types (65.5%) types (34.5%)
y w y w w m w m

wild type white wild type miniature

F2
y w y w males w m w m

yellow, white yellow white, miniature white

y w y w w m w m

y w y w w m w m
wild type white wild type miniature
F2
y w y w females w m w m

y w y w w m w m
yellow, white yellow white, miniature white

FIGURE 73 Data representing the F1 and F2 results of crosses involving the yellow body (y), white eye (w) mutations (cross A),
and the white eye, miniature wing (m) mutations (cross B), as compiled by Sturtevant. In cross A, 0.5 percent of the F2 flies (males
and females) demonstrate recombinant phenotypes, which express either white or yellow. In cross B, 34.5 percent of the F2 flies
(males and females) demonstrate recombinant phenotypes, which express either miniature or white.

low. However, both of these genes are much farther apart from
miniature gene because the whiteminiature and
yellowminiature combinations show larger recombination
frequencies. Because miniature shows more recombination
with yellow than with white (35.4% versus 34.5%), it follows
that white is located between the other two genes, not outside
of them.
Sturtevant knew from Morgans work that the frequency of
exchange could be used as an estimate of the distance
between two genes or loci along the chromosome. He con-
structed a chromosome map of the three genes on the X
chromosome, setting 1 map unit (mu) equal to 1 percent
recombination between two genes.* In the preceding exam-
ple, the distance between yellow and white is thus 0.5 mu,
and the distance between yellow and miniature is 35.4 mu. It
follows that the distance between white and miniature
should be 35.4 - 0 .5 = 34.9 mu. This estimate is close to
the actual frequency of recombination between white and
miniature (34.5%). The map for these three genes is shown
*In honor of Morgans work, 1 map unit is now referred to as a cen-
timorgan (cM).
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 142

142 Chapter 7 Linkage and Chromosome Mapping in Eukaryotes

0.5 34.5 of inheritance was still widely disputedeven Morgan was

y w m Research has now firmly established that chromosomes con-
35.4
FIGURE 74 A map of the yellow body (y), white eye (w), and
miniature wing (m) genes on the X chromosome of Drosophila
melanogaster. Each number represents the percentage of
recombinant offspring produced in one of three crosses, each
involving two different genes.
in Figure 74. The fact that these do not add up perfectly is
due to the imprecision of independently-conducted map-
ping experiments.
In addition to these three genes, Sturtevant considered two
other genes on the X chromosome and produced a more exten-
sive map that included all five genes. He and a colleague,
Calvin Bridges, soon began a search for autosomal linkage in
Drosophila. By 1923, they had clearly shown that linkage and
crossing over are not restricted to X-linked genes but can also
be demonstrated with autosomes. During this work, they made
another interesting observation. Crossing over in Drosophila
was shown to occur only in females. The fact that no crossing
over occurs in males made genetic mapping much less com-
plex to analyze in Drosophila. However, crossing over does
occur in both sexes in most other organisms.
Although many refinements in chromosome mapping have
been developed since Sturtevants initial work, his basic prin-
ciples are considered to be correct. These principles are used
to produce detailed chromosome maps of organisms for
which large numbers of linked mutant genes are known.
Sturtevants findings are also historically significant to the
broader field of genetics. In 1910, the chromosomal theory
skeptical of this theory before he conducted his experiments.
tain genes in a linear order and that these genes are the equiv-
alent of Mendels unit factors.
Single Crossovers
Why should the relative distance between two loci influence
the amount of recombination and crossing over observed
between them? During meiosis, a limited number of crossover
events occur in each tetrad. These recombinant events occur
randomly along the length of the tetrad. Therefore, the closer
two loci reside along the axis of the chromosome, the less
likely any single crossover event will occur between them.
The same reasoning suggests that the farther apart two linked
loci, the more likely a random crossover event will occur
between them.
In Figure 75(a), a single crossover occurs between two
nonsister chromatids but not between the two loci; therefore,
the crossover is not detected because no recombinant gametes
are produced. In Figure 75(b), where two loci are quite far
apart, crossover does occur between them, yielding recombi-
nant gametes.
When a single crossover occurs between two nonsister chro-
matids, the other two chromatids of the tetrad are not involved
in this exchange and enter the gamete unchanged. Even if a
single crossover occurs 100 percent of the time between two
linked genes, recombination is subsequently observed in only
50 percent of the potential gametes formed. This concept is
diagrammed in Figure 76. Theoretically, if we consider only
single exchanges and observe 20 percent recombinant
gametes, crossing over actually occurred in 40 percent of the

(a) FIGURE 75 Two examples of a single crossover

Exchange Gametes between two nonsister chromatids and the
A B A B gametes subsequently produced. In (a), the
exchange does not alter the linkage arrangement
A B
between the alleles of the two genes, only parental
A B Meiosis
a b gametes form, and the exchange goes undetected.
a b In (b), the exchange separates the alleles and
a b results in recombinant gametes, which are
a b detectable.
Segments of two ...but the linkage between
nonsister chromatids the A and B alleles and
are exchanged... between the a and b
alleles is unchanged.

(b)
Exchange Gametes
A B A B

A B
a b
a b
Segments of two
nonsister chromatids
are exchanged...
A b
Meiosis
a B
a b
...and alleles have
recombined in two of
the four gametes.
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 143

7.3 Determining the Gene Sequence During Mapping Relies on the Analysis of Multiple Crossovers 143

A B

Three-Point Mapping
Web Tutorial 7.3
A B
a b

a b

A B A b a B a b

Noncrossover Crossover Crossover Noncrossover

gamete gamete gamete gamete

FIGURE 76 The consequences of a single exchange between two nonsister chromatids occurring in the tetrad stage.
Two noncrossover (parental) and two crossover (recombinant) gametes are produced.

tetrads. Under these conditions, the general rule is that the
percentage of tetrads involved in an exchange between two
genes is twice the percentage of recombinant gametes pro-
duced. Therefore, the theoretical limit of observed recombi-
nation due to crossing over is 50 percent.
When two linked genes are more than 50 mu apart, a
crossover can theoretically be expected to occur between
them in 100 percent of the tetrads. If this prediction were
achieved, each tetrad would yield equal proportions of the
four gametes shown in Figure 76, just as if the genes were on
different chromosomes and assorting independently. However,
this theoretical limit is seldom achieved.
7.3 Determining the Gene Sequence
During Mapping Relies on the
Analysis of Multiple Crossovers
The study of single crossovers between two linked genes pro-
vides the basis of determining the distance between them.
However, when many linked genes are studied, their sequence
along the chromosome is more difficult to determine. Fortu-
nately, the discovery that multiple exchanges occur between
the chromatids of a tetrad has facilitated the process of pro-
ducing more extensive chromosome maps. As we shall see
next, when three or more linked genes are investigated simul-
taneously, it is possible to determine first the sequence of the
genes and then the distances between them.
Multiple Crossovers
It is possible that in a single tetrad, two, three, or more
exchanges will occur between nonsister chromatids as a
result of several crossover events. Double exchanges of
genetic material result from double crossovers (DCOs), as
shown in Figure 77. For a double exchange to be studied,
three gene pairs must be investigated, each heterozygous for
two alleles. Before we determine the frequency of recombi-
nation among all three loci, lets review some simple proba-
bility calculations.
As we have seen, the probability of a single exchange occur-
ring between the A and B or the B and C genes relates directly
to the distance between the respective loci. The closer A is to
B and B is to C, the less likely a single exchange will occur
between either of the two sets of loci. In the case of a double
crossover, two separate and independent events or exchanges
must occur simultaneously. The mathematical probability of

A B C

A B C
a b c

a b c
Chiasmata

A b C a B c

Double-crossover gametes

A B C a b c

Noncrossover gametes

FIGURE 77 Consequences of a double exchange between two nonsister chromatids. Because the exchanges involve only two
chromatids, two noncrossover gametes and two double-crossover gametes are produced. The photograph shows several chiasmata found
in a tetrad isolated during the first meiotic prophase. See also the chapter opening photograph on p.136.
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 144
144 Chapter 7 Linkage and Chromosome Mapping in Eukaryotes
two independent events occurring simultaneously is equal to
the product of the individual probabilities (the product law).
Suppose that crossover gametes resulting from single
exchanges are recovered 20 percent of the time (p = 0.20)
between A and B, and 30 percent of the time (p = 0.30)
between B and C. The probability of recovering a double-
crossover gamete arising from two exchanges (between A and
B, and between B and C) is predicted to be
(0.20)(0.30) = 0.06, or 6 percent. It is apparent from this
calculation that the frequency of double-crossover gametes is
always expected to be much lower than that of either single-
crossover class of gametes.
If three genes are relatively close together along one chromo-
some, the expected frequency of double-crossover gametes is
extremely low. For example, suppose the AB distance in
Figure 77 is 3 mu and the BC distance is 2 mu. The expect-
ed double-crossover frequency is (0.03)(0.02) = 0.0006, or
0.06 percent. This translates to only 6 events in 10,000. Thus,
in a mapping experiment where closely linked genes are
involved, very large numbers of offspring are required to
detect double-crossover events. In this example, it is unlikely
that a double crossover will be observed even if 1000 offspring
are examined. Thus, it is evident that if four or five genes are
being mapped, even fewer triple and quadruple crossovers can
be expected to occur.
Now Solve This
Problem 14 on page 161 asks you to contrast the results
of crossing over when the arrangement of alleles along
the homologs differs in an organism heterozygous for
three genes.
Hint: Homologs enter noncrossover gametes
unchanged, and all crossover results must be derived
from the noncrossover sequence of alleles.
Three-Point Mapping in Drosophila
The information in the preceding section enables us to map
three or more linked genes in a single cross. To illustrate the
mapping process in its entirety, we examine two situations
involving three linked genes in two quite different organisms.
To execute a successful mapping cross, three criteria must
be met:
1. The genotype of the organism producing the
crossover gametes must be heterozygous at all loci
under consideration.
2. The cross must be constructed so that genotypes of all
gametes can be determined accurately by observing the
phenotypes of the resulting offspring. This is necessary
because the gametes and their genotypes can never be
observed directly. To overcome this problem, each phe-
notypic class must reflect the genotype of the gametes of
the parents producing it.
3. A sufficient number of offspring must be produced in the
mapping experiment to recover a representative sample
of all crossover classes.
These criteria are met in the three-point mapping cross from
D. melanogaster shown in Figure 78. In this cross, three
X-linked recessive mutant genesyellow body color (y),
white eye color (w), and echinus eye shape (ec)are consid-
ered. To diagram the cross, we must assume some theoretical
sequence, even though we do not yet know if it is correct. In
Figure 78, we initially assume the sequence of the three
genes to be ywec. If this is incorrect, our analysis will
demonstrate this and reveal the correct sequence.
In the P1 generation, males hemizygous for all three wild-
type alleles are crossed to females that are homozygous for all
three recessive mutant alleles. Therefore, the P1 males are
wild type with respect to body color, eye color, and eye shape.
They are said to have a wild-type phenotype. The females, on
the other hand, exhibit the three mutant traitsyellow body
color, white eyes, and echinus eye shape.
This cross produces an F1 generation consisting of females
that are heterozygous at all three loci and males that, because
of the Y chromosome, are hemizygous for the three mutant
alleles. Phenotypically, all F1 females are wild type, while all
F1 males are yellow, white, and echinus. The genotype of the
F1 females fulfills the first criterion for mapping the three
linked genes; that is, it is heterozygous at the three loci and
can serve as the source of recombinant gametes generated by
crossing over. Note that because of the genotypes of the P1
parents, all three mutant alleles are on one homolog and all
three wild-type alleles are on the other homolog. Other
arrangements are possible. For example, the heterozygous F1
female might have the y and ec mutant alleles on one homolog
and the w allele on the other. This would occur if, in the P1
cross, one parent was yellow, echinus and the other parent
was white.
In our cross, the second criterion is met by virtue of the
gametes formed by the F1 males. Every gamete contains
either an X chromosome bearing the three mutant alleles or
a Y chromosome, which is genetically inert for the three
loci being considered. Whichever type participates in fertil-
ization, the genotype of the gamete produced by the F1
female will be expressed phenotypically in the F2 male and
female offspring derived from it. Thus, all F1 noncrossover
and crossover gametes can be detected by observing the F2
phenotypes.
With these two criteria met, we can now construct a chromo-
some map from the crosses shown in Figure 78. First, we
determine which F2 phenotypes correspond to the various
noncrossover and crossover categories. To determine the non-
crossover F2 phenotypes, we must identify individuals
derived from the parental gametes formed by the F1 female.
Each such gamete contains an X chromosome unaffected by
crossing over. As a result of segregation, approximately equal
proportions of the two types of gametes and, subsequently,
the F2 phenotypes, are produced. Because they derive from a
heterozygote, the genotypes of the two parental gametes and
the resultant F2 phenotypes complement one another. For
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 145

7.3 Determining the Gene Sequence During Mapping Relies on the Analysis of Multiple Crossovers 145

y w ec y+ w+ ec+
P1

y w ec

Gametes y w ec y+ w+ ec+

y w ec y w ec
F1
y+ w+ ec+

Gametes

y w ec Category,
Origin of F2 Observed total, and
female gametes Gametes phenotype Number percentage
1 y w ec y w ec
y w ec
NCO y w ec 4685 Non-
y w ec crossover
y w ec
2 y w ec y w ec
y w ec
y w ec y w ec 4759 9444
94.44%
y w ec
SCO 3 y w ec y w ec
y w ec Single
y w ec 80
y w ec crossover
y w ec between y
and w
4 y w ec y w ec
y w ec 150
y w ec y w ec 70
1.50%
y w ec

SCO 5 y w ec y w ec
y w ec y w ec 193 Single
y w ec crossover
y w ec between w
and ec
6 y w ec y w ec
y w ec 207 400
y w ec y w ec
4.00%
y w ec

DCO
7 y w ec y w ec Double
y w ec
y w ec 3 crossover
y w ec between y
y w ec and w
and between
8 y w ec y w ec w and ec
y w ec
y w ec y w ec 3 6
y w ec 0.06%

y w ec
Map of y, w, and ec loci
1.56 4.06

FIGURE 78 A three-point mapping cross involving the yellow (y or y + ), white (w or w + ), and echinus (ec or ec + ) genes in D.
melanogaster. NCO, SCO, and DCO refer to noncrossover, single-crossover, and double-crossover groups, respectively.
Centromeres are not included on the chromosomes, and only two nonsister chromatids are shown initially.
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 146
146 Chapter 7 Linkage and Chromosome Mapping in Eukaryotes
example, if one is wild type, the other is completely mutant.
This is the case in the cross being considered. In other situa-
tions, if one chromosome shows one mutant allele, the second
chromosome shows the other two mutant alleles, and so on.
They are therefore called reciprocal classes of gametes and
phenotypes.
The two noncrossover phenotypes are most easily recog-
nized because they exist in the greatest proportion. Figure 78
shows that gametes 1 and 2 are present in the greatest num-
bers. Therefore, flies that express yellow, white, and echinus
phenotypes and flies that are normal (or wild type) for all
three characters constitute the noncrossover category and rep-
resent 94.44 percent of the F2 offspring.
The second category that can be easily detected is repre-
sented by the double-crossover phenotypes. Because of
their low probability of occurrence, they must be present in
the least numbers. Remember that this group represents
two independent but simultaneous single-crossover events.
Two reciprocal phenotypes can be identified: gamete 7,
which shows the mutant traits yellow, echinus but normal
eye color; and gamete 8, which shows the mutant trait
white but normal body color and eye shape. Together these
double-crossover phenotypes constitute only 0.06 percent
of the F2 offspring.
The remaining four phenotypic classes represent two cate-
gories resulting from single crossovers. Gametes 3 and 4, re-
ciprocal phenotypes produced by single-crossover events
occurring between the yellow and white loci, are equal to
1.50 percent of the F2 offspring; gametes 5 and 6, constituting
4.00 percent of the F2 offspring, represent the reciprocal phe-
notypes resulting from single-crossover events occurring
between the white and echinus loci.
The map distances separating the three loci can now be
calculated. The distance between y and w or between w and
ec is equal to the percentage of all detectable exchanges
occurring between them. For any two genes under consider-
ation, this includes all appropriate single crossovers as well
as all double crossovers. The latter are included because
they represent two simultaneous single crossovers. For the y
and w genes, this includes gametes 3, 4, 7, and 8, totaling
1.50% + 0.06% , or 1.56 mu. Similarly, the distance
between w and ec is equal to the percentage of offspring
resulting from an exchange between these two loci:
gametes 5, 6, 7, and 8, totaling 4.00% + 0 .06% , or 4.06
mu. The map of these three loci on the X chromosome is
shown at the bottom of Figure 78.
Determining the Gene Sequence
In the preceding example, the sequence (or order) of the
three genes along the chromosome was assumed to be
ywec. Our analysis shows this sequence to be consistent
with the data. However, in most mapping experiments the
gene sequence is not known, and this constitutes another
variable in the analysis. In our example, had the gene
sequence been unknown, it could have been determined
using a straightforward method.
This method is based on the fact that there are only three
possible arrangements, each containing one of the three genes
between the other two:
(I) wyec (y in the middle)
(II) yecw (ec in the middle)
(III) ywec (w in the middle)
Use the following steps during your analysis to determine
the gene order:
1. Assuming any one of the three orders, first determine the
arrangement of alleles along each homolog of the het-
erozygous parent giving rise to noncrossover and
crossover gametes (the F1 female in our example).
2. Determine whether a double-crossover event occurring
within that arrangement will produce the observed double-
crossover phenotypes. Remember that these phenotypes
occur least frequently and are easily identified.
3. If this order does not produce the predicted phenotypes,
try each of the other two orders. One must work!
These steps are shown in Figure 79, using our ywec cross.
The three possible arrangements are labeled I, II, and III, as
shown above.
1. Assuming that y is between w and ec, arrangement I of
alleles along the homologs of the F1 heterozygote is
w y ec
w + y + ec +
We know this because of the way in which the P1 genera-
tion was crossed: The P1 female contributes an X chro-
mosome bearing the w, y, and ec alleles, while the P1
male contributes an X chromosome bearing the w + , y + ,
and ec + alleles.
2. A double crossover within that arrangement yields the
following gametes:
w y+ ec and w+ y ec +
Following fertilization, if y is in the middle, the F2
double-crossover phenotypes will correspond to these
gametic genotypes, yielding offspring that express the
white, echinus phenotype and offspring that express
the yellow phenotype. Instead, however, determination
of the actual double-crossover phenotypes reveals them
to be yellow, echinus flies and white flies. Therefore, our
assumed order is incorrect.
3. If we consider arrangement II with the ec/ec + alleles in
the middle or arrangement III with the w/w + alleles in the
middle:
y ec w y w ec
(II) or (III)
y + ec + w + y + w + ec +
we see that arrangement II again provides predicted dou-
ble-crossover phenotypes that do not correspond to the
actual (observed) double-crossover phenotypes. The pre-
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 147

7.3 Determining the Gene Sequence During Mapping Relies on the Analysis of Multiple Crossovers 147

Three theoretical sequences Double-crossover gametes Phenotypes

w y ec w y ec
white, echinus
I
yellow
w y ec w y ec

y ec w y ec w
yellow, white
II
echinus
y ec w y ec w

y w ec y w ec
yellow, echinus
III
white
y w ec y w ec

FIGURE 79 The three possible sequences of the white, yellow, and echinus genes, the results of a double crossover in each case, and
the resulting phenotypes produced in a testcross. For simplicity, the two noncrossover chromatids of each tetrad are omitted.

dicted phenotypes are yellow, white flies and echinus flies
in the F2 generation. Therefore, this order is also incorrect.
However, arrangement III produces the observed pheno-
typesyellow, echinus flies and white flies. Therefore,
this arrangement, with the w gene in the middle, is correct.
To summarize, this method is rather straightforward: First
determine the arrangement of alleles on the homologs of the het-
erozygote yielding the crossover gametes by locating the recip-
rocal noncrossover phenotypes. Then, test each of three possible
orders to determine which yields the observed double-crossover
phenotypesthe one that does so represents the correct order.
A Mapping Problem in Maize
Having established the basic principles of chromosome map-
ping, we now consider a related problem in maize (corn), in
which the gene sequence and interlocus distances are unknown.
This analysis differs from the preceding example in two ways.
First, the previous mapping cross involved X-linked genes.
Here, we consider autosomal genes. Second, in the discussion
of this cross we have changed our use of symbols, as first sug-
gested in Chapter 4. Instead of using the gene symbols and
superscripts (e.g., bm+ , v + , and pr +), we simply use + to
denote each wild-type allele. This system is easier to manipu-
late but requires a better understanding of mapping procedures.
When we look at three autosomally linked genes in maize, the
experimental cross must still meet the same three criteria we
established for the X-linked genes in Drosophila: (1) One parent
must be heterozygous for all traits under consideration; (2) the
gametic genotypes produced by the heterozygote must be appar-
ent from observing the phenotypes of the offspring; and (3) a
sufficient sample size must be available for complete analysis.
In maize, the recessive mutant genes brown midrib (bm),
virescent seedling (v), and purple aleurone (pr) are linked on
chromosome 5. Assume that a female plant is known to be
heterozygous for all three traits, but we do not know (1) the
arrangement of the mutant alleles on the maternal and pater-
nal homologs of this heterozygote, (2) the sequence of genes,
or (3) the map distances between the genes. What genotype
must the male plant have to allow successful mapping? To
meet the second criterion, the male must be homozygous for
all three recessive mutant alleles. Otherwise, offspring of this
cross showing a given phenotype might represent more than
one genotype, making accurate mapping impossible.
Figure 710 diagrams this cross. As shown, we know neither
the arrangement of alleles nor the sequence of loci in the het-
erozygous female. Several possibilities are shown, but we
have yet to determine which is correct. We dont know the
sequence in the testcross male parent either, and so we must
designate it randomly. Note that we have initially placed v in
the middle. This may or may not be correct.
The offspring are arranged in groups of two for each pair of
reciprocal phenotypic classes. The two members of each reci-
procal class are derived from no crossing over (NCO), one of
two possible single-crossover events (SCO), or a double
crossover (DCO).
To solve this problem, refer to Figures 710 and 711 as you
consider the following questions.
1. What is the correct heterozygous arrangement of alleles in
the female parent? Determine the two noncrossover class-
es, those that occur with the highest frequency. In this case,
they are + v bm and pr+ + . Therefore, the alleles on the
homologs of the female parent must be arranged as shown
in Figure 711(a). These homologs segregate into gametes,
unaffected by any recombination event. Any other arrange-
ment of alleles will not yield the observed noncrossover
classes. (Remember that + v bm is equivalent to pr + v bm,
and that pr + + is equivalent to pr v + bm+).
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 148

148 Chapter 7 Linkage and Chromosome Mapping in Eukaryotes

(a) Some possible allele arrangements and gene sequences in a heterozygous female

v  bm v pr bm v  pr v  

 pr      bm   bm pr

pr v bm pr  bm v bm 

    v    pr

? ? ? pr v bm

Which of the above

is correct? ? ? ? pr v bm
Heterozygous Testcross
female male

(b) Actual results of mapping cross*

Phenotypes Number Total and Exchange

of offspring percentage classification

 v bm 230
467 Noncrossover
42.1% (NCO)
pr   237

  bm 82
161 Single crossover
14.5% (SCO)
pr v  79

 v  200
395 Single crossover
35.6% (SCO)
pr  bm 195

pr v bm 44
86 Double crossover
7.8% (DCO)
   42

* The sequence pr v bm may or may not be correct.

FIGURE 710 (a) Some possible allele arrangements and gene sequences in a heterozygous female. The data from a three-point mapping cross
depicted in (b), where the female is testcrossed, determine which combination of arrangement and sequence is correct [see Figure 711(d)].

2. What is the correct sequence of genes? We know that the + bm v v + bm

arrangement of alleles is or
pr + + + pr +
+ v bm
Only the order on the right yields the observed double-
pr + + crossover gametes [Figure 711(d)]. Therefore, the pr
gene is in the middle. From this point on, work the prob-
But is the gene sequence correct? That is, will a double-
lem using this arrangement and sequence, with the pr
crossover event yield the observed double-crossover
locus in the middle.
phenotypes after fertilization? Observation shows that it
will not [Figure 711(b)]. Now try the other two orders 3. What is the distance between each pair of genes? Having
[Figures 711(c) and (d)] maintaining the same established the sequence of loci as vprbm, we can
arrangement: determine the distance between v and pr and between pr
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 149

7.3 Determining the Gene Sequence During Mapping Relies on the Analysis of Multiple Crossovers 149

Allele arrangement Testcross Explanation

and sequence phenotypes

(a)  v bm
 v bm Noncrossover phenotypes provide the basis
of determining the correct arrangement of
and
alleles on homologs
pr  
pr  

(b)
 v bm
  bm Expected double-crossover phenotypes
and if v is in the middle
pr v 
pr  
(c)
 bm v   v Expected double-crossover phenotypes
if bm is in the middle
and
pr bm 
pr  

(d)
v  bm v pr bm Expected double-crossover phenotypes
and if pr is in the middle
(This is the actual situation.)
  
 pr 
(e)
v  bm v pr  Given that (a) and (d) are correct, single-
crossover phenotypes when exchange
and
occurs between v and pr
  bm
 pr 

(f)
v  bm
v   Given that (a) and (d) are correct, single-
and crossover phenotypes when exchange
occurs between pr and bm
 pr bm
 pr 

(g) v pr bm
Final map
22.3 43.4

FIGURE 711 Producing a map of the three genes in the testcross of Figure 710, where neither the arrangement of alleles nor the
sequence of genes in the heterozygous female parent is known.

and bm. Remember that the map distance between two

genes is calculated on the basis of all detectable recombi-
nation events occurring between them. This includes
both single- and double-crossover events.
Figure 711(e) shows that the phenotypes v pr + and + +bm
result from single crossovers between the v and pr loci, account-
ing for 14.5 percent of the offspring [according to data in Figure
10(b)]. By adding the percentage of double crossovers (7.8%) to
the number obtained for single crossovers, the total distance
between the v and pr loci is calculated to be 22.3 mu.
Figure 711(f) shows that the phenotypes v + + and
+ pr bm result from single crossovers between the pr and bm
loci, totaling 35.6 percent. Added to the double crossovers
(7.8%), the distance between pr and bm is calculated to be
43.4 mu. The final map for all three genes in this example is
shown in Figure 711(g).
Now Solve This
Problem 20 on page 162 asks you to solve a three-point
mapping problem where only six phenotypic categories
are observed, when eight categories are typical of such
a cross.
Hint: If the distances between each pair of genes is
relatively small, reciprocal pairs of single crossovers
will be evident, but the sample size may be too small
to recover the predicted number of double crossovers,
excluding their appearance. You should write the
missing gametes down, assign them as double
crossovers, and record zeroes for their frequency of
appearance.
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 150
150 Chapter 7 Linkage and Chromosome Mapping in Eukaryotes
7.4 As the Distance Between Two
Genes Increases, Mapping
Estimates Become More Inaccurate
Virtual Crossover Laboratory
So far, we have assumed that crossover frequencies are directly pro-
portional to the distance between any two loci along the chromo-
some. However, it is not always possible to detect all crossover
Web Tutorial 7.4
events. A case in point is a double exchange that occurs between
the two loci in question. As shown in Figure 712(a), if a double
exchange occurs, the original arrangement of alleles on each non-
sister homolog is recovered. Therefore, even though crossing over
has occurred, it is impossible to detect. This phenomenon is true
for all even-numbered exchanges between two loci.
Furthermore, as a result of complications posed by multiple-
strand exchanges, mapping determinations usually underesti-
mate the actual distance between two genes. The farther apart
two genes are, the greater the probability that undetected
crossovers will occur. While the discrepancy is minimal for two
genes relatively close together, the degree of inaccuracy increas-
es as the distance increases, as shown in the graph of recombi-
nation frequency versus map distance in Figure 712(b). There,
the theoretical frequency where a direct correlation between
recombination and map distance exists is contrasted with the
actual frequency observed as the distance between two genes
increases. The most accurate maps are constructed from experi-
ments where genes are relatively close together.
Interference and the Coefficient of Coincidence
As shown in our maize example, we can predict the expected
frequency of multiple exchanges, such as double crossovers,
once the distance between genes is established. For example,
(a) Two-strand double exchange
A B
A B
a b
a b
(b)
50
% Recombinant chromatids
40
(Hypothetical)
30
(Actual)
20
10
10 20 30 40 50 60
Map distance (map units)
in the maize cross, the distance between v and pr is 22.3 mu,
and the distance between pr and bm is 43.4 mu. If the two sin-
gle crossovers that make up a double crossover occur inde-
pendently of one another, we can calculate the expected
frequency of double crossovers 1DCOexp2:
DCOexp = (0.223) * (0.434) = 0.097 = 9.7%
Often in mapping experiments, the observed DCO frequency is
less than the expected number of DCOs. In the maize cross, for
example, only 7.8 percent DCOs are observed when 9.7 percent
are expected. Interference (I), the phenomenon where a
crossover event in one region of the chromosome inhibits a sec-
ond event in nearby regions, causes this reduction.
To quantify the disparities that result from interference, we
calculate the coefficient of coincidence (C):
Observed DCO
C =
Expected DCO
In the maize cross, we have
0.078
C = = 0.804
0.097
Once we have found C, we can quantify interference using
the simple equation
I = 1 - C
In the maize cross, we have
I = 1.000 - 0.804 = 0.196
If interference is complete and no double crossovers occur,
then I = 1 .0 . If fewer DCOs than expected occur, I is a posi-
A B
A B
No detectable
a b recombinants
a b
FIGURE 712 (a) A double crossover is undetected
because no rearrangement of alleles occurs. (b) The
theoretical and actual percentage of recombinant
chromatids versus map distance. The straight line shows
the theoretical relationship if a direct correlation
between recombination and map distance exists. The
70 80 curved line is the actual relationship derived from
studies of Drosophila, Neurospora, and Zea mays.
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 151

7.5 Drosophila Genes Have Been Extensively Mapped 151

tive number and positive interference has occurred. If more
DCOs than expected occur, I is a negative number and nega-
tive interference has occurred. In the maize example, I is a
positive number (0.196), indicating that 19.6 percent fewer
double crossovers occurred than expected.
Positive interference is most often observed in eukaryotic
systems. In general, the closer genes are to one another along
the chromosome, the more positive interference occurs. In
fact, interference in Drosophila is often complete within a dis-
tance of 10 mu, and no multiple crossovers are recovered.
This observation suggests that physical constraints preventing
the formation of closely aligned chiasmata contribute to inter-
ference. This interpretation is consistent with the finding that
interference decreases as the genes in question are located far-
ther apart. In the maize cross in Figures 710 and 711, the
three genes are relatively far apart and 80 percent of the
expected double crossovers are observed.
7.5 Drosophila Genes Have Been
Extensively Mapped
In organisms such as Drosophila, maize, and the mouse,
where large numbers of mutations have been discovered and
where mapping crosses are possible, extensive chromosome
maps have been constructed. Figure 713 shows partial maps
for the four chromosomes (I, II, III, and IV) of Drosophila.

I (X) II III IV

yellow body, y
0.0 scute bristles, sc 0.0 aristaless antenna, al 0.0 roughoid eyes, ru 0.0 cubitus interruptus
0.2 veinlet veins, ve veins, ci
1.3 Star eyes, S 1.4 Roughened eye, R 0.2 grooveless scutellum, gvl
1.5 white eyes, w
1.4 bent wings, bt
3.0 facet eyes, fa
5.5 echinus eyes, ec 2.0 eyeless, ey
6.1 Curly wing, Cy
7.5 ruby eyes, rb shaven bristles, sv
13.0 dumpy wings, dp 11.0 female sterile, 3.0
crossveinless sparkling eyes, spa
13.7 fs(3)G2
wings, cv 16.5 clot eyes, cl 17.0 raisin eye, rai
20.0 cut wings, ct 19.2 javelin bristles, jv
22.0 Sternopleural
21.0 singed bristles, sn bristles, Sp 26.0 sepia eyes, se
27.5 tan body, t 26.5 hairy body, h
27.7 lozenge eyes, lz
31.0 dachs tarsus, d
33.0 vermilion eyes, v 36.0 corrugated, corr 35.5 eyes gone, eyg
36.1 miniature 39.3 daughterless, da
wings, m 40.5 Lyra wings, Ly
41.0 Jammed wings, J 41.0 Dichaete bristles, D
43.0 sable body, s 43.2 thread arista, th
44.0 scarlet eyes, st
44.0 garnet eyes, g 48.5 black body, b
51.0 reduced bristles, rd 50.0 curled wings, cu
51.5 scalloped
wings, sd
54.5 purple eyes, pr
56.7 forked bristles, f
57.5 cinnabar eyes, cn 58.2 Stubble bristles, Sb
57.0 Bar eyes, B
58.5 spineless brisles, ss
59.5 fused veins, fu
61.0 withered wing, whd
62.5 carnation eyes, car 62.0 stripe body, sr
66.0 bobbed hairs, bb 66.2 Delta veins, DI
67.0 vestigial wings, vg
68.1 little fly, lf 69.5 Hairless bristles, H
72.0 Lobe eyes, L 70.7 ebony body, e
75.5 curved wings, c 74.7 cardinal eyes, cd
77.5 obtuse wing, obt
83.1 adipose, adp FIGURE 713 A partial
genetic map of the four
90.0 disrupted wing, dsr 88.0 mahogany eyes, mah
91.1 rough eyes, ro
chromosomes of D.
91.5 smooth
melanogaster. The circle on
abdomen, sm 95.5 suppression of purple, su-pr
100.5 plexus wings, px each chromosome represents
100.7 claret eyes, ca
104.0 heavy wing veins, hv the position of the centromere.
104.5 brown eyes, bw 106.2 Minute bristles, M(3)g Chromosome I is the X
107.0 speck body, sp chromosome. Chromosome IV
is not drawn to scale.
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 152
152 Chapter 7 Linkage and Chromosome Mapping in Eukaryotes
Virtually every morphological feature of the fruit fly has been
subjected to mutations. The locus of each mutant gene is first
localized to one of the four chromosomes (or linkage groups)
and then mapped in relation to all other genes present on that
chromosome. Based on cytological evidence, the relative
lengths of these genetic maps correlate roughly with the rela-
tive physical lengths of these chromosomes.
7.6 Lod Score Analysis and Somatic
Cell Hybridization Were
Historically Important in Creating
Human Chromosome Maps
In humans, where neither designed matings nor large numbers
of offspring are available, the earliest linkage studies were
based on pedigree analysis. Attempts were made to establish
whether a trait was X-linked or autosomal. For autosomal
traits, geneticists tried to distinguish whether pairs of traits
demonstrate linkage or independent assortment. In this way, it
was hoped that a human gene map could be created.
The difficulty arises, however, when two genes of interest are
separated on a chromosome such that recombinant gametes are
formed, obscuring linkage in a pedigree. In such cases, the
demonstration of linkage is enhanced by an approach that relies
on probability, called the lod score method. First devised by J.
B. S. Haldane and C. A. Smith in 1947, and refined by Newton
Morton in 1955, the lod score (log of the odds favoring linkage)
assesses the probability that a particular pedigree involving two
traits reflects linkage or not. First, the probability is calculated
that family data (pedigrees) concerning two or more traits con-
form to the transmission of traits without linkage. Then the
probability is calculated that the identical family data following
these same traits result from linkage with a specified recombi-
nation frequency. The ratio of these probabilities expresses the
odds for and against linkage.
Accuracy using the lod score method is limited by the extent
of the family data, but nevertheless represents an important
advance in assigning human genes to specific chromosomes
and constructing preliminary human chromosome maps. The
initial results were discouraging because of the methods
inherent limitations and the relatively high haploid number of
Hybrid
cell Human chromosomes present
lines
1 2 3 4 5 6
23
34
41
FIGURE 714 A hypothetical grid of data used in synteny testing to assign genes to their appropriate human chromosomes.
Three somatic hybrid-cell lines, designated 23, 34, and 41, have each been scored for the presence or absence of human
chromosomes 18, as well as for their ability to produce the hypothetical human gene products A, B, C, and D.
human chromosomes (23), and by 1960, almost no linkage or
mapping information had become available.
In the 1960s, a new technique, somatic cell hybridization,
proved to be an immense aid in assigning human genes to their
respective chromosomes. This technique, first discovered by
Georges Barsky, relies on the fact that two cells in culture can
be induced to fuse into a single hybrid cell. Barsky used two
mouse-cell lines, but it soon became evident that cells from dif-
ferent organisms could also be fused. When fusion occurs, an
initial cell type called a heterokaryon is produced. The hybrid
cell contains two nuclei in a common cytoplasm. By using the
proper techniques, it is possible to fuse human and mouse cells,
for example, and isolate the hybrids from the parental cells.
As the heterokaryons are cultured in vitro, two interesting
changes occur. The nuclei eventually fuse, creating a
synkaryon. Then, as culturing is continued for many genera-
tions, chromosomes from one of the two parental species are
gradually lost. In the case of the humanmouse hybrid,
human chromosomes are lost randomly until eventually the
synkaryon has a full complement of mouse chromosomes and
only a few human chromosomes. As we shall see, it is the
preferential loss of human chromosomes (rather than mouse
chromosomes) that makes possible the assignment of human
genes to the chromosomes on which they reside.
The experimental rationale is straightforward. If a specific
human gene product is synthesized in a synkaryon containing
one to three human chromosomes, then the gene responsible
for that product must reside on one of the human chromo-
somes remaining in the hybrid cell. On the other hand, if the
human gene product is not synthesized in a synkaryon, the
gene responsible is not present on those human chromosomes
that remain in this hybrid cell. Ideally, a panel of 23 hybrid-
cell lines, each with just one unique human chromosome,
would allow the immediate assignment of any human gene for
which the product could be characterized.
In practice, a panel of cell lines, each with several remaining
human chromosomes, is most often used. The correlation of the
presence or absence of each chromosome with the presence or
absence of each gene product is called synteny testing. Con-
sider, for example, the hypothetical data provided in Figure
714, where four gene products (A, B, C, and D) are tested in
relationship to eight human chromosomes. Lets analyze the
gene that produces product A.
Gene products
expressed
7 8 A B C D
+ +
+ +
+ + +
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 153

7.6 Lod Score Analysis and Somatic Cell Hybridization Were Historically Important in Creating Human Chromosome Maps 153

1. Product A is not produced by cell line 23, but chromosomes
1, 2, 3, and 4 are present in cell line 23. Therefore, we rule
out the presence of gene A on those four chromosomes and
conclude that it must be on chromosome 5, 6, 7, or 8.
2. Product A is produced by cell line 34, which contains
chromosomes 5 and 6 but not 7 and 8. Therefore, gene A
is on chromosome 5 or 6, but cannot be on 7 or 8 because
they are absent even though product A is produced.
3. Product A is also produced by cell line 41, which con-
tains chromosome 5 but not chromosome 6. Using sim-
ilar reasoning we see that gene A must be on
chromosome 5.
Using a similar approach, we can assign gene B to chromo-
some 3. Perform this analysis for yourself to demonstrate that
this is correct.
Gene C presents a unique situation. The data indicate that it
is not present on chromosomes 17. While it might be on
chromosome 8, no direct evidence supports this conclusion,
and other panels are needed. We leave gene D for you to
analyzeon what chromosome does it reside?
Using this technique, researchers have assigned literally
hundreds of human genes to one chromosome or another.
Some of the assignments shown in Figure 715 were either
derived or confirmed with the use of somatic cell hybridiza-
tion techniques. To map genes for which the products have yet

Chromosome 1

sn 6 GDH
5
4
3
3 Rh
X Chromosome 2
2 1
2 p
1
p DMD
2
2 1 AMY 1,2
1 2
3 PGM -1
1
1 1
1 1
1
2
1 2 1
PGK
3
2
q 1 3 AT3
q 2
2 4
2 3 4
5 5
6 HGPRT
7 1
8 CB
HEMA 3 PEPC
G6PD 2
1 2 FHM
4 3
4

Key AMY Amylase (salivary and pancreatic)

AT3 Antithrombin (clotting factor IV)
CB Color Blindness
DMD Duchenne Muscular Dystrophy
FHM Fumarate Hydratase (mitochondrial)
GDH Glucose Dehydrogenase
G6PD Glucose-6-Phosphate Dehydrogenase
HEMA Hemophilia A (classic)
HGPRT Hypoxanthine-Guanine-Phosphoribosyl
Transferase (LeschNyhan syndrome)
PEPC Peptidase C
FIGURE 715 Representative regional gene assignments for PGK Phosphoglycerate Kinase
human chromosome 1 and the X chromosome. Many assignments PGM Phosphoglucomutase
were initially derived using somatic cell hybridization techniques. Rh Rhesus Blood Group (erythroblastosis fetalis)
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 154
154 Chapter 7 Linkage and Chromosome Mapping in Eukaryotes
to be discovered, researchers have had to rely on other
approaches. For example, by using recombinant DNA tech-
nology in conjunction with pedigree analysis, it has been pos-
sible to assign the genes responsible for Huntington disease,
cystic fibrosis, and neurofibromatosis to their respective
chromosomes, 4, 7, and 17.
7.7 Linkage and Mapping Studies
Can Be Performed in Haploid
Organisms
Many of the single-celled eukaryotes are haploid during the
vegetative stages of their life cycle. The alga Chlamydomonas
and the mold Neurospora demonstrate this genetic condition.
But these organisms do form reproductive cells that fuse dur-
ing fertilization, producing a diploid zygote. However, this
structure soon undergoes meiosis, resulting in haploid vegeta-
tive cells that then propagate by mitotic divisions. In genetic
studies, small haploid organisms have several important
advantages compared with diploid eukaryotes. They can be
cultured and manipulated in genetic crosses much more easily.
In addition, a haploid organism contains only a single allele of
each gene, which is expressed directly in the phenotype. This
greatly simplifies genetic analysis. As a result, organisms
such as Chlamydomonas and Neurospora serve as the sub-
jects of research investigations in many areas of genetics,
including linkage and mapping studies.
In order to perform genetics experiments with such organ-
isms, crosses are made, and following fertilization, the mei-
otic structures may be isolated. Because all four meiotic
products give rise to spores, the structures bearing these
products (asci) are called tetrads. The term tetrad has a dif-
ferent meaning here than earlier when it was used to
describe a precise chromatid configuration in meiosis. Indi-
vidual tetrads are isolated, and the resultant cells are grown
and analyzed separately from those of other tetrads. In the
results we are about to describe, the data reflect the propor-
tion of tetrads that show one combination of genotypes, the
proportion that show another combination, and so on. Such
experimentation is called tetrad analysis.
Gene-to-Centromere Mapping
When a single gene in Neurospora is analyzed (Figure 716),
the data can be used to calculate the map distance between the
gene and the centromere. This process is sometimes referred
to as mapping the centromere. It is accomplished by experi-
mentally determining the frequency of recombination using
tetrad data. Note that once the four meiotic products of the
tetrad are formed, a mitotic division occurs, resulting in eight
ordered products (ascospores). If no crossover event occurs
between the gene under study and the centromere, the pattern
of ascospores contained within an ascus (pl., asci) appears as
shown in Figure 716(a), aaaa+ + + + .*
*The pattern (+ + + +aaaa) can also be formed. However, it is indis-
tinguishable from (aaaa+ + + +).
This pattern represents first-division segregation because
the two alleles are separated during the first meiotic division.
However, a crossover event will alter this pattern, as shown
in Figure 716(b), aa+ +aa+ + , and Figure 716(c),
+ +aaaa+ + . Two other recombinant patterns occur but are
not shown: + +aa+ +aa and aa+ + + +aa. These depend on
the chromatid orientation during the second meiotic division.
These four patterns reflect second-division segregation
because the two alleles are not separated until the second mei-
otic division. Usually, the ordered tetrad data are condensed to
reflect the genotypes of the identical ascospore pairs, and six
combinations are possible.
First-Division Second-Division
Segregation Segregation
aa+ + a+a+ +aa+
+ +aa +a+a a+ +a
To calculate the distance between the gene and the cen-
tromere, a large number of asci resulting from a controlled
cross are counted. We then use these data to calculate the dis-
tance (d):
1>2 1second-division segregant asci2
d =
total asci scored
The distance (d) reflects the percentage of recombination and
is only half the number of second-division segregant asci.
This is because crossing over in each occurs in only two of the
four chromatids during meiosis.
To illustrate, we use a for albino and + for wild type in
Neurospora. In crosses between the two genetic types, suppose
we observe 65 first-division segregants, and 70 second-division
segregants. The distance between a and the centromere is
(1>2) (70)
d = = 0.259
135
or about 26 mu.
As the distance increases to 50 mu, all asci should reflect
second-division segregation. However, numerous factors pre-
vent this. As in diploid organisms, mapping accuracy based
on crossover events is greatest when the gene and centromere
are relatively close together.
We can also analyze haploid organisms to distinguish
between linkage and independent assortment of two genes
mapping distances between gene loci are calculated once
linkage is established. As a result, detailed maps of organisms
such as Neurospora and Chlamydomonas are now available.
7.8 Other Aspects of Genetic
Exchange
Careful analysis of crossing over during gamete formation
allows us to construct chromosome maps in both diploid and
haploid organisms. However, we should not lose sight of the
real biological significance of crossing over, which is to gen-
erate genetic variation in gametes and, subsequently, in the
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 155

7.8 Other Aspects of Genetic Exchange 155

FIGURE 716 Three ways in which different ascospore patterns

can be generated in Neurospora. Analysis of these patterns is the
basis of gene-to-centromere mapping. The photograph shows a
variety of ascospore arrangements within Neurospora asci.

Condition Four-strand Chromosomes Chromosomes Ascospores

stage following following in ascus
meiosis mitotic division
a
(a) a a
a a
a
a a
a a
No a
a
crossover   
  
  
  

First-division segregation

a a
(b) a
a
a
a  

One form of 
a 
crossover in  a a
a
four-strand a
stage a
 
 



Second-division segregation

(c)  


a a
An alternate a a
a
crossover in a a
four-strand  a
a a
stage a
a
 
 



Second-division segregation

offspring derived from the resultant eggs and sperm. Because
of the critical role of crossing over in generating variation,
the study of genetic exchange is a key topic for study in
genetics. But many important questions remain. For example,
does crossing over involve an actual exchange of chromo-
some arms? Does exchange occur between paired sister chro-
matids during mitosis? We shall briefly consider possible
answers to these questions.
Cytological Evidence for Crossing Over
Once genetic mapping was understood, it was of great interest
to investigate the relationship between chiasmata observed in
meiotic prophase I and crossing over. For example, are chias-
mata visible manifestations of crossover events? If so, then
crossing over in higher organisms appears to result from an
actual physical exchange between homologous chromo-
somes. That this is the case was demonstrated independently
in the 1930s by Harriet Creighton and Barbara McClintock in
Zea mays, and by Curt Stern in Drosophila.
Since the experiments are similar, we will consider only the
work with maize. Creighton and McClintock studied two
linked genes on chromosome 9. At one locus, the alleles
colorless (c) and colored (C) control endosperm coloration.
At the other locus, the alleles starchy (Wx) and waxy (wx)
control the carbohydrate characteristics of the endosperm.
The maize plant studied is heterozygous at both loci. The key
to this experiment is that one of the homologs contains two
unique cytological markers. The markers consist of a densely
stained knob at one end of the chromosome and a translocated
piece of another chromosome (8) at the other end. The
arrangements of alleles and cytological markers can be
detected cytologically and are shown in Figure 717.
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 156

156 Chapter 7 Linkage and Chromosome Mapping in Eukaryotes

Parents Recombinant offspring

translocated segment
Case l Case ll
knob
C wx c Wx c wx C Wx

c Wx c wx c wx c wx
Colored, starchy Colorless, starchy Colorless, waxy Colored, starchy

FIGURE 717 The phenotypes and chromosome compositions of parents and recombinant offspring in Creighton and McClintocks
experiment in maize. The knob and translocated segment are the cytological markers that established that crossing over involves an
actual exchange of chromosome arms.

Creighton and McClintock crossed this plant to a plant
homozygous for the colored allele (c) and heterozygous for
the endosperm alleles. They obtained a variety of different
phenotypes in the offspring, but they were most interested
in a crossover result involving the chromosome with the
unique cytological markers. They examined the chromo-
somes of this plant with the colorless, waxy phenotype
(case I in Figure 717) for the presence of the cytological
markers. If physical exchange between homologs accompa-
nies genetic crossing over, the translocated chromosome
will still be present, but the knob will notthis is exactly
what happened. In a second plant (case II), the phenotype
colored, starchy should result from either nonrecombinant
gametes or from crossing over. Some of the plants then
ought to contain chromosomes with the dense knob but not
the translocated chromosome. This condition was also
found, and the conclusion that a physical exchange takes
place was again supported. Along with Sterns findings
with Drosophila, this work clearly established that crossing
over has a cytological basis.
replication, each pair of sister chromatids has one member
with one strand of DNA labeled with BUdR and the other
member with both strands labeled with BUdR. Using a dif-
ferential stain, chromatids with the analog in both strands
stain less brightly than chromatids with BUdR in only one
strand. As a result, SCEs are readily detectable if they occur.
In Figure 718, numerous instances of SCE events are
clearly evident. These sister chromatids are sometimes
referred to as harlequin chromosomes because of their
patchlike appearance.
While the significance of SCEs is still uncertain, several
observations have led to great interest in this phenomenon.
We know, for example, that agents that induce chromosome
damage (viruses, X-rays, ultraviolet light, and certain chem-

Sister Chromatid Exchanges

Knowing that crossing over occurs between synapsed
homologs in meiosis, we might ask whether a similar physical
exchange occurs between homologs during mitosis. While
homologous chromosomes do not usually pair up or synapse
in somatic cells (Drosophila is an exception), each individual
chromosome in prophase and metaphase of mitosis consists
of two identical sister chromatids, joined at a common cen-
tromere. Surprisingly, several experimental approaches have
demonstrated that reciprocal exchanges similar to crossing
over occur between sister chromatids. These sister chro-
matid exchanges (SCEs) do not produce new allelic combi-
nations, but evidence is accumulating that attaches
significance to these events. FIGURE 718 Light micrograph of sister chromatid exchanges
Identification and study of SCEs are facilitated by several (SCEs) in mitotic chromosomes. Sometimes called harlequin
modern staining techniques. In one technique, cells replicate chromosomes because of their patchlike appearance, chromatids
for two generations in the presence of the thymidine analog with the thymidine analog BUdR in both DNA strands fluoresce
bromodeoxyuridine (BUdR).* Following two rounds of less brightly than do those with the analog in only one strand.
These chromosomes were stained with 33258-Hoechst reagent
and acridine orange and then viewed using fluorescence
*The abbreviation BrdU is also used to denote bromodeoxyuridine. microscopy.
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 157

7.9 Did Mendel Encounter Linkage? 157

ical mutagens) increase the frequency of SCEs. The frequen-

cy of SCEs is also elevated in Bloom syndrome, a human
disorder caused by a mutation in the BLM gene on chromo-
some 15. This rare, recessively inherited disease is charac-
terized by prenatal and postnatal retardation of growth, a
great sensitivity of the facial skin to the sun, immune defi-
ciency, a predisposition to malignant and benign tumors, and
abnormal behavior patterns. The chromosomes from cul-
tured leukocytes, bone marrow cells, and fibroblasts derived
from homozygotes are very fragile and unstable compared
to those of homozygous and heterozygous normal individu-
als. Increased breaks and rearrangements between nonho-
mologous chromosomes are observed in addition to
excessive amounts of sister chromatid exchanges. Work by
James German and colleagues suggests that the BLM gene
encodes an enzyme called DNA helicase, which is best
known for its role in DNA replication.
7.9 Did Mendel Encounter Linkage?
We conclude this chapter by examining a modern-day inter-
pretation of the experiments that form the cornerstone of
transmission geneticsMendels crosses with garden peas.
Some observers believe that Mendel had extremely good
fortune in his classic experiments. He did not encounter
apparent linkage relationships between the seven mutant
characters in any of his crosses. Had Mendel obtained highly
variable data characteristic of linkage and crossing over, these
unorthodox ratios might have hindered his successful analysis
and interpretation.
The article by Stig Blixt, reprinted in its entirety in the fol-
lowing box, demonstrates the inadequacy of this hypothesis.
As we shall see, some of Mendels genes were indeed linked.
We leave it to Stig Blixt to enlighten you as to why Mendel
did not detect linkage.

Why Didnt Gregor Mendel Find Linkage?

It is quite often said that Mendel was very
fortunate not to run into the complication
of linkage during his experiments. He
used seven genes, and the pea has only
seven chromosomes. Some have said that
had he taken just one more, he would have
had problems. This, however, is a gross
oversimplification. The actual situation,
most probably, is shown in Table 7.1. This
shows that Mendel worked with three
genes in chromosome 4, two genes in
chromosome 1, and one gene in each of
chromosomes 5 and 7. It seems at first
glance that, out of the 21 dihybrid combi-
nations Mendel theoretically could have
studied, no fewer than four (that is, ai,
vfa, vle, fale) ought to have resulted in
linkages. However, as found in hundreds
of crosses and shown by the genetic map
of the pea, a and i in chromosome 1 are so
distantly located on the chromosome that
no linkage is normally detected. The same
is true for v and le on the one hand, and fa
on the other, in chromosome 4. This
leaves vle, which ought to have shown
linkage.
Mendel, however, seems not to have
published this particular combination and
thus, presumably, never made the appro-
priate cross to obtain both genes segregat-
ing simultaneously. It is therefore not so
astonishing that Mendel did not run into
the complication of linkage, although he
did not avoid it by choosing one gene
from each chromosome.
STIG BLIXT
Weibullsholm Plant Breeding Institute,
Landskrona, Sweden,
and Centro Energia Nucleate na
Agricultura, Piracicaba, SP, Brazil.
Source: Reprinted by permission from Nature,
Vol. 256, p. 206. 1975 Macmillan Magazines
Limited.

TABLE 7.1 Relationship between modern genetic terminology and character pairs used by Mendel

Character Pair Used by Mendel Alleles in Modern Terminology Located in Chromosome

Seed color, yellowgreen Ii 1

Seed coat and flowers, coloredwhite Aa 1
Mature pods, smooth expandedwrinkled indented Vv 4
Inflorescences, from leaf axisumbellate in top of plant Fafa 4
Plant height, 0.51 m Lele 4
Unripe pods, greenyellow Gpgp 5
Mature seeds, smoothwrinkled Rr 7
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 158

158 Chapter 7 Linkage and Chromosome Mapping in Eukaryotes

CHAPTER SUMMARY
1. Genes located on the same chromosome are said to be linked. Al-
leles located on the same homolog, therefore, may be transmitted
together during gamete formation. However, crossing over be-
tween homologs during meiosis results in the reshuffling of alleles
and thereby contributes to genetic variability within gametes.
2. Early in the twentieth century, geneticists realized that crossing
over provides an experimental basis for mapping the location of
linked genes relative to one another along the chromosome.
3. Somatic cell hybridization techniques have made possible link-
age and mapping analysis of human genes.
4. Mapping studies may also be performed with haploid organisms
such as Chlamydomonas and Neurospora.
5. Cytological investigations of both maize and Drosophila reveal
that crossing over involves a physical exchange of segments
between nonsister chromatids.
6. An exchange of genetic material between sister chromatids
can occur during mitosis as well. These events are referred
to as sister chromatid exchanges (SCEs). An elevated fre-
quency of such events is seen in the human disorder Bloom
syndrome.
7. Evidence now suggests that several of the genes studied by
Mendel are, in fact, linked. However, in such cases, the genes are
sufficiently far apart to prevent the detection of linkage.

KEY TERMS
Bloom syndrome, 157 harlequin chromosomes, 156 noncrossover gametes, 137, 140
bromodeoxyuridine (BUdR), 156 heterokaryon, 152 parental gametes, 137
chiasmata, 140 Huntington disease, 154 reciprocal classes, 146
chromosomal theory of inheritance, 142 interference (I), 150 recombinant gametes, 137, 140
chromosome map, 141 linkage, 137 second-division segregation, 154
coefficient of coincidence (C), 150 linkage group, 140 sister chromatid exchanges (SCEs), 156
complete linkage, 000 linkage ratio, 138 somatic cell hybridization, 152
crossing over, 137 lod score method, 152 synkaryon, 152
crossover gametes, 137 mapping the centromere, 154 synteny testing, 152
cystic fibrosis, 154 multiple-strand exchange, 150 tetrad, 154
double crossover (DCO), 143 neurofibromatosis, 154 tetrad analysis, 154
first-division segregation, 154

Insights and Solutions

1. In rabbits, black color (B) is dominant to brown (b), while full
B C b c ch
color (C) is dominant to chinchilla (cch). The genes controlling
these traits are linked. Rabbits that are heterozygous for both

traits and express black, full color are crossed to rabbits that
express brown, chinchilla with the following results: b c ch b c ch
31 brown, chinchilla
Noncrossovers
34 black, full B C b c ch
16 brown, full
19 black, chinchilla
Determine the arrangement of alleles in the heterozygous parents b c ch b c ch
and the map distance between the two genes. black, full brown, chinchilla

Solution: This is a two-point map problem, where the two most

Single crossovers
prevalent reciprocal phenotypes are the noncrossovers. The less B c ch b C
frequent reciprocal phenotypes arise from a single crossover. The
arrangement of alleles is derived from the noncrossover pheno-
types because they enter gametes intact.
The single crossovers give rise to 35/100 offspring (35%). b c ch b c ch
black, chincilla brown, full
Therefore, the distance between the two genes is 35 mu.
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 159

Insights and Solutions 159

2. In Drosophila, Lyra (Ly) and Stubble (Sb) are dominant muta- Case A
tions located at locus 40 and 58, respectively, on chromosome III. br Ly Sb
A recessive mutation with bright red eyes is discovered and
shown also to be located on chromosome III. A map is obtained
by crossing a female who is heterozygous for all three mutations
to a male that is homozygous for the bright-red mutation (which   
we will call br), and the data in the table are generated. Deter-
mine the location of the br mutation on chromosome III.
Double crossovers
Phenotype Number br  Sb
(1) Ly Sb br 404
and
(2) + + + 422
(3) Ly + + 18  Ly 
(4) + Sb br 16
(5) Ly + br 75 Case B
Ly br Sb
(6) + Sb + 59
(7) Ly Sb + 4 and
(8) + + br 2
Total = 1000   

Solution: First, determine the arrangement of the alleles on the

homologs of the heterozygous crossover parent (the female in
Double crossovers
this case). To do this, locate the most frequent reciprocal pheno-
types, which arise from the noncrossover gametesthese are Ly  Sb
phenotypes (1) and (2). Each phenotype represents the arrange-
and
ment of alleles on one of the homologs. Therefore, the arrange-
ment is  
br
Ly Sb br
Comparison with the actual data shows that case B is correct.
The double-crossover gametes yield flies that express Ly and Sb
but not br, or express br but not Ly and Sb. Therefore, the correct
   arrangement and sequence are shown below.

Second, find the correct sequence of the three loci along the Ly br Sb
chromosome. This is done by determining which sequence yields
the observed double-crossover phenotypes, which are the least
frequent reciprocal phenotypes (7 and 8). If the sequence is cor-   
rect as written, then the double crossover depicted here,
Once this sequence is found, determine the location of br relative
Ly Sb br
to Ly and Sb. A single crossover between Ly and br, as shown here,

Ly br Sb
  

will yield Ly + br and + Sb + as phenotypes. Inspection   

shows that these categories (5 and 6) are actually single
crossovers, not double crossovers. Therefore, the sequence is
incorrect, as written. Only two other sequences are possible:
The br gene is either to the left of Ly (case A), or it is between
Ly and Sb (case B).
yields flies that are Ly + + and + br Sb (phenotypes 3 and 4).
Therefore, the distance between the Ly and br loci is equal to
(18 + 16 + 4 + 2)/1000 = 40/1000 = 0.04 = 4 mu

(Cont. on the next page)

KLUGMC07_136-163hr 10/23/06 4:44 PM Page 160

160 Chapter 7 Linkage and Chromosome Mapping in Eukaryotes

Remember to add the double crossovers because they represent
two single crossovers occurring simultaneously. You need to
know the frequency of all crossovers between Ly and br, so they
must be included.
Similarly, the distance between the br and Sb loci is derived
mainly from single crossovers between them.
Ly br Sb
   tions are alleles, no complementation will occur and all progeny
This event yields Ly br + and + + Sb phenotypes (phenotypes 5
and 6). Therefore, the distance equals
(75 + 59 + 4 + 2)/1000 = 140/1000 = 0.14 = 14 mu
The final map shows that br is located at locus 44, since Lyra and
Stubble are known.
3. Refer to Figure 713, and predict what gene (which we called
br) was discovered on chromosome III in Problem 2. Suggest an
experimental cross that could confirm your prediction.
Solution: Inspection of Figure 713 reveals that the mutation
scarlet (st) is present at locus 44.0, so it is reasonable to hypothe-
size that the bright-red eye mutation is an allele at the scarlet
locus.
To test this hypothesis, you could perform complementation
analysis (see Chapter 4) by crossing females expressing the
bright-red mutation with known scarlet males. If the two muta-
will reveal a bright-red mutant eye phenotype. If complementa-
tion occurs, all progeny will express normal brick-red (wild-
type) eyes, since the bright-red mutation and scarlet are at
different loci (they are probably very close together). In such a
case, all progeny will be heterozygous at both the bright-red eye
and the scarlet loci and will not express either mutation because
they are both recessive. This type of complementation analysis is
called an allelism test.

40 (4) 44 (14) 58

Ly br Sb

PROBLEMS AND DISCUSSION QUESTIONS

1. What is the significance of crossing over (which leads to genetic
recombination) to the process of evolution?
2. Describe the cytological observation that suggests that crossing
over occurs during the first meiotic prophase.
3. Why does more crossing over occur between two distantly
linked genes than between two genes that are very close togeth-
er on the same chromosome?
4. Why is a 50 percent recovery of single-crossover products the
upper limit, even when crossing over always occurs between two
linked genes?
5. Why are double-crossover events expected less frequently than
single-crossover events?
6. What is the proposed basis for positive interference?
7. What three essential criteria must be met in order to execute a
successful mapping cross?
8. The genes dumpy wings (dp), clot eyes (cl), and apterous wings
(ap) are linked on chromosome II of Drosophila. In a series of
two-point mapping crosses, the genetic distances shown below
were determined. What is the sequence of the three genes?
(c) a and b are linked on the same autosome but are so close
together that a crossover almost never occurs.
(d) a and b are linked on the same autosome about 10 mu apart.
10. Colored aleurone in the kernels of corn is due to the dominant
allele R. The recessive allele r, when homozygous, produces col-
orless aleurone. The plant color (not kernel color) is controlled
by another gene with two alleles, Y and y. The dominant Y allele
results in green color, whereas the homozygous presence of the
recessive y allele causes the plant to appear yellow. In a testcross
between a plant of unknown genotype and phenotype and a plant
that is homozygous recessive for both traits, the following prog-
eny were obtained:
colored, green 88
colored, yellow 12
colorless, green 8
colorless, yellow 92

dpap 42 Explain how these results were obtained by determining the

exact genotype and phenotype of the unknown plant, including
dpcl 3 the precise association of the two genes on the homologs (i. e.,
apcl 39 the arrangement).
11. In the cross shown here, involving two linked genes, ebony (e)
and claret (ca), in Drosophila, where crossing over does not
9. Consider two hypothetical recessive autosomal genes a and b,
occur in males, offspring were produced in a (2 + :1 ca:1 e) phe-
where a heterozygote is testcrossed to a double-homozygous
notypic ratio:
mutant. Predict the phenotypic ratios under the following
O P
conditions:
(a) a and b are located on separate autosomes. e ca + e ca +
(b) a and b are linked on the same autosome but are so far apart *
e+ ca e+ ca
that a crossover always occurs between them.
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 161

Problems and Discussion Questions 161

These genes are 30 mu apart on chromosome III. What did termination of sex was made in the F2 data. (a) Using proper
crossing over in the female contribute to these phenotypes? nomenclature, determine the genotypes of the P1 and F1 parents.
12. With two pairs of genes involved (P, p and Z, z), a testcross (to (b) Determine the sequence of the three genes and the map dis-
ppzz) with an organism of unknown genotype indicated that the tance between them. (c) Are there more or fewer double
gametes were produced in these proportions: PZ = 42.4%; crossovers than expected? (d) Calculate the coefficient of coinci-
Pz = 6 .9%; pZ = 7 .1%; and pz = 43.6%. Draw all possible dence; does this represent positive or negative interference?
conclusions from these data.
13. In a series of two-point map crosses involving five genes located Phenotype Offspring
on chromosome II in Drosophila, the following recombinant
(single-crossover) frequencies were observed: sc s v 314
+ + + 280
pradp 29
+ s v 150
prvg 13
sc + + 156
prc 21
sc + v 46
prb 6
+ s + 30
adpb 35
sc s + 10
adpc 8
+ + v 14
adpvg 16
vgb 19 16. A cross in Drosophila involved the recessive, X-linked genes
yellow body (y), white eyes (w), and cut wings (ct). A yellow-
vgc 8 bodied, white-eyed female with normal wings was crossed to a
cb 27 male whose eyes and body were normal, but whose wings were
cut. The F1 females were wild type for all three traits, while the
(a) If the adp gene is present near the end of chromosome II F1 males expressed the yellow-body, white-eye traits. The cross
(locus 83), construct a map of these genes. was carried to F2 progeny, and only male offspring were tallied.
(b) In another set of experiments, a sixth gene (d) was tested On the basis of the data shown here, a genetic map was con-
against b and pr, and the results were d b = 17% and structed. (a) Diagram the genotypes of the F1 parents. (b) Con-
dpr = 23%. Predict the results of two-point maps between struct a map, assuming that w is at locus 1.5 on the X
d and c, d and vg, and d and adp. chromosome. (c) Were any double-crossover offspring expect-
14. Two different female Drosophila were isolated, each heterozy- ed? (d) Could the F2 female offspring be used to construct the
gous for the autosomally linked genes black body (b), dachs map? Why or why not?
tarsus (d), and curved wings (c). These genes are in the order
dbc, with b closer to d than to c. Shown below is the genotypic Phenotype Male Offspring
arrangement for each female, along with the various gametes y + ct 9
formed by both. Identify which categories are noncrossovers
(NCO), single crossovers (SCO), and double crossovers (DCO) + w + 6
in each case. Then, indicate the relative frequency in which each y w ct 90
will be produced.
+ + + 95
Female A Female B + + ct 424
d b + d + +
+ + c + b c y w + 376
:
:

Gamete formation y + + 0
+ w ct 0
(1) d b c (5) d + + (1) d b + (5) d b c
(2) + + + (6) + b c (2) + + c (6) + + + 17. In Drosophila, Dichaete (D) is a mutation on chromosome III
(3) + + c (7) d + c (3) d + c (7) d + + with a dominant effect on wing shape. It is lethal when ho-
mozygous. The genes ebony body (e) and pink eye (p) are re-
(4) d b + (8) + b + (4) + b + (8) + b c cessive mutations on chromosome III. Flies from a Dichaete
stock were crossed to homozygous ebony, pink flies, and the
15. In Drosophila, a cross was made between females expressing the F1 progeny with a Dichaete phenotype were backcrossed to
three X-linked recessive traits, scute bristles (sc), sable body (s), the ebony, pink homozygotes. Using the results of this back-
and vermilion eyes (v) and wild-type males. All females were cross shown in the table, (a) diagram the cross, showing the
wild type in the F1 , while all males expressed all three mutant genotypes of the parents and offspring of both crosses. (b)
traits. The cross was carried to the F2 generation, and 1000 off- What is the sequence and interlocus distance between these
spring were counted, with the results shown in the table. No de- three genes?
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 162

162 Chapter 7 Linkage and Chromosome Mapping in Eukaryotes

Phenotype Number AaBbCc 20 AaBbcc 20

Dichaete 401 aabbCc 20 aabbcc 20
ebony, pink 389 AabbCc 5 Aabbcc 5
Dichaete, ebony 84 aaBbCc 5 aaBbcc 5
pink 96
(a) Assuming simple dominance and recessiveness in each gene
Dichaete, pink 2 pair, if these three genes were all assorting independently,
how many genotypic and phenotypic classes would result in
ebony 3
the offspring, and in what proportion?
Dichaete, ebony, pink 12 (b) Answer part (a) again, assuming the three genes are so tightly
linked on a single chromosome that no crossover gametes
wild type 13 were recovered in the sample of offspring.
(c) What can you conclude from the actual data about the loca-
18. Drosophila females homozygous for the third chromosomal tion of the three genes in relation to one another?
genes pink eye (p) and ebony body (e) were crossed with males 24. Based on our discussion of the potential inaccuracy of mapping
homozygous for the second chromosomal gene dumpy wings (see Figure 712), would you revise your answer to Problem
(dp). Because these genes are recessive, all offspring were wild 23(c)? If so, how?
type (normal). F1 females were testcrossed to triply recessive 25. In a plant, fruit color is either red or yellow, and fruit shape is
males. If we assume that the two linked genes (p and e) are 20 mu either oval or long. Red and oval are the dominant traits. Two
apart, predict the results of this cross. If the reciprocal cross were plants, both heterozygous for these traits, were testcrossed, with
made (F1 maleswhere no crossing over occurswith triply the results shown below. Determine the location of the genes rel-
recessive females), how would the results vary, if at all? ative to one another and the genotypes of the two parental plants.
19. In Drosophila, the two mutations Stubble bristles (Sb) and
curled wings (cu) are linked on chromosome III. Sb is a domi- Progeny
nant gene that is lethal in a homozygous state, and cu is a reces-
sive gene. If a female of the genotype Phenotype Plant A Plant B
Sb cu red, long 46 4

+ + yellow, oval 44 6
is to be mated to detect recombinants among her offspring, what red, oval 5 43
male genotype would you choose as her mate?
yellow, long 5 47
20. In Drosophila, a heterozygous female for the X-linked recessive
traits a, b, and c was crossed to a male that phenotypically Total 100 100
expressed a, b, and c. The offspring occurred in the phenotypic
ratios in the following table, and no other phenotypes were 26. In a plant heterozygous for two gene pairs (Ab/aB) , where the
observed. (a) What is the genotypic arrangement of the alleles of two loci are linked and 25 mu apart, two such individuals were
these genes on the X chromosome of the female? (b) Determine crossed together. Assuming that crossing over occurs during
the correct sequence, and construct a map of these genes on the the formation of both male and female gametes and that the A
X chromosome. (c) What progeny phenotypes are missing, and and B alleles are dominant, determine the phenotypic ratio of
why? the offspring.
27. In a cross in Neurospora involving two alleles, B and b, the
+ b c 460 tetrad patterns in the following table were observed. Calculate
the distance between the gene and the centromere.
a + + 450
a b c 32 Tetrad Pattern Number
+ + + 38 BBbb 36
a + c 11 bbBB 44
+ b + 9 BbBb 4
bBbB 6
21. Are sister chromatid exchanges effective in producing genetic
variability in an individual? in the offspring of individuals? BbbB 3
22. What conclusion can be drawn from the observations that in
male Drosophila, no crossing over occurs and that during meio- bBBb 7
sis, synaptonemal complexes (ultrastructural components found
between synapsed homologs in meiosis) are not seen in males 28. In Creighton and McClintocks experiment demonstrating that
but are observed in females where crossing over occurs? crossing over involves physical exchange between chromosomes
23. An organism of the genotype AaBbCc was testcrossed to a triply (see Section 7.8), explain the importance of the cytological
recessive organism (aabbcc). The genotypes of the progeny are markers (the translocated segment and the chromosome knob) in
in the following table. the experimental rationale.
KLUGMC07_136-163hr 10/23/06 4:44 PM Page 163

Problems and Discussion Questions 163

29. A number of humanmouse somatic cell hybrid clones were ex- rated the mutant into a stock containing the recessive gene black
amined for the expression of specific human genes and the pres- (b, body color, located on chromosome II) and the recessive
ence of human chromosomesthe results are summarized in gene pink (p, eye color, located on chromosome III). A female
this table. Assign each gene to the chromosome upon which it is from the homozygous black, pink, short stock was then mated to
located. a wild-type male. The F1 males of this cross were all wild type
and were then backcrossed to the homozygous b, p, sh females.
Hybrid-Cell Clone The F2 results appeared as shown in the table that follows, and
no other phenotypes were observed. (a) Based on these results,
A B C D E F the student was able to assign sh to a linkage group (a chromo-
Genes (expressed or not) some). Determine which chromosome, and include step-by-step
ENO1 (enolase-1) - + - + + - reasoning. (b) The student repeated the experiment, making the
reciprocal cross: F1 females backcrossed to homozygous b, p, sh
MDH1 (malate dehydrogenase-1) + + - + - + males. She observed that 85 percent of the offspring fell into the
PEPS (peptidase S) + - + - - - given classes, but that 15 percent of the offspring were equally
divided among b+p , b+ + , +shp, and +sh+ phenotypic males
PGM1 (phosphoglucomutase-1) - + - + + - and females. How can these results be explained, and what
information can be derived from these data?
Chromosomes (present or absent)
1 - + - + + - Phenotype Female Male
2 + + - + - + wild 63 59
3 + + - - + - pink* 58 65
4 + - + - - - black, short 55 51
5 - + + + + + black, pink, short 69 60

30. A female of genotype *Pink indicates that the other two traits are wild type (normal).
Similarly, black, short offspring are wild type for eye color.
a b c
+ + + 32. In Drosophila, a female fly is heterozygous for three mutations,
produces 100 meiotic tetrads. Of these, 68 show no crossover Bar eyes (B), miniature wings (m), and ebony body (e). (Note
events. Of the remaining 32, 20 show a crossover between a and that Bar is a dominant mutation.) The fly is crossed to a male
b, 10 show a crossover between b and c, and 2 show a double with normal eyes, miniature wings, and ebony body. The results
crossover between a and b and between b and c. Of the 400 of the cross are shown below. Interpret the results of this cross. If
gametes produced, how many of each of the 8 different geno- you conclude that linkage is involved between any of the genes,
types will be produced? Assuming the order abc and the allele determine the map distance(s) between them.
arrangement shown above, what is the map distance between
these loci? miniature 111 Bar 117
31. D. melanogaster has one pair of sex chromosomes (XX or XY) wild 29 Bar, miniature 26
and three autosomes (chromosomes II, III, and IV). A genetics
student discovered a male fly with very short (sh) legs. Using Bar, ebony 101 ebony 35
this male, the student was able to establish a pure-breeding stock Bar, miniature, ebony 31 miniature, ebony 115
of this mutant and found that it was recessive. She then incorpo-
KLUGMC16_357-375v2 9/1/06 2:14 PM Page 357

C H A P T E R

16
Cell-Cycle Regulation
and Cancer

Colored scanning electron micro-

graph of two prostate cancer cells
in the final stages of cell division
(cytokinesis). The cells are still
joined by strands of cytoplasm.

CHAPTER CONCEPTS
Cancer is a group of genetic diseases
affecting fundamental aspects of cellular
function, including DNA repair, the cell cycle,
apoptosis, differentiation, cell migration and
cellcell contact.
Most cancer-causing mutations occur in
somatic cells; only about 1 percent of
cancers have a hereditary component.
Mutations in cancer-related genes lead to
abnormal proliferation and loss of control
over how cells spread and invade
surrounding tissues.
The development of cancer is a multistep
process requiring mutations in genes
controlling many aspects of cell
proliferation and metastasis.
Cancer cells show high levels of genomic
instability, leading to the accumulation of
multiple mutations in cancer-related
genes.
Mutations in proto-oncogenes and tumor
suppressor genes contribute to the
development of cancers.
Oncogenic viruses introduce oncogenes into
infected cells and stimulate cell
proliferation.
Environmental agents contribute to cancer
by damaging DNA.

357
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 358

358 Chapter 16 Cell-Cycle Regulation and Cancer

C
ancer is the second leading cause of death in Western
countries, surpassed only by heart disease. It strikes
people of all ages, and one out of three people will
experience a cancer diagnosis sometime in his or her life-
time. Each year, more than 1 million cases of cancer are
diagnosed in the United States and more than 500,000 peo-
ple die from the disease (Table 16.1).
Over the last 30 years, scientists have discovered that can-
cer is a genetic disease, characterized by an interplay of
mutant forms of oncogenes and tumor suppressor genes
leading to the uncontrolled growth and spread of cancer
cells. While some of these mutations may be inherited, as
we will see, most all of them that lead to disease occur in
somatic cells that then divide and lead to tumors. Comple-
tion of the Human Genome Project is opening the door to a
wealth of new information about the mutations that trigger
a cell to become cancerous. This new understanding of can-
cer genetics is also leading to new gene-specific treatments,
some of which are now entering clinical trials. Some scien-
tists predict that gene therapies will replace chemotherapies
within the next 25 years.
The goal of this chapter is to highlight our current under-
standing of the nature and causes of cancer. As we will see,
cancer is a genetic disease that arises from mutations in
genes controlling many basic aspects of cellular function.
We will examine the relationship between genes and can-
cer, and consider how mutations, chromosomal changes,
and environmental agents play roles in the development of
cancer.
16.1 Cancer Is a Genetic Disease at the
Level of Somatic Cells
Perhaps the most significant development in understanding the
causes of cancer is the realization that cancer is a genetic disease.
Genomic alterations that are associated with cancer range from
single-nucleotide substitutions to large-scale chromosome
rearrangements, amplifications, and deletions (Figure 161).
However, unlike other genetic diseases, cancer is caused by muta-
tions that occur predominantly in somatic cells. Only about 1 per-
cent of cancers are associated with germ-line mutations that
increase a persons susceptibility to certain types of cancer.
Another important difference between cancer and other genetic
diseases is that cancers rarely arise from a single mutation, but
from the accumulation of many mutationsas many as six to ten.
The mutations that lead to cancer affect multiple cellular func-
tions, including repair of DNA damage, cell division, apoptosis,
cellular differentiation, migratory behavior, and cellcell contact.
What Is Cancer?
Clinically, cancer is defined as a large number of complex dis-
eases, up to a hundred, that behave differently depending on the
cell types from which they originate. Cancers vary in their ages
of onset, growth rates, invasiveness, prognoses, and responsive-
(a)

1 2 3 4 5

How Do We Know?
In this chapter, we will focus on cancer as a genetic dis- 6 7 8 9 10 11 12
ease, with an emphasis on the relationship between can-
cer and DNA damage, as well as on the multiple genetic
13 14 15 16 17 18
steps that lead to cancer. We also discuss how cancer cells
show defects in cell-cycle regulation. We conclude with an
investigation of the roles played by viruses and environ- x y
19 20 21 22
mental agents in the development of cancer. As you study
this topic, you should try to answer several fundamental (b)
questions:
1. How do we know that cancers arise from a single cell that
contains mutations? 1 2 3 4 5
2. How do we know that cancer development requires more
than one mutation?
6 7 8 9 10 11 12
3. How do we know that cancer cells contain defects in DNA
repair? 13 14 15 16 17 18
4. How do we know that most cancers are not
hereditary? 19 20 21 22 x

5. How do we know that some viruses contain genes that FIGURE 161 (a) Spectral karyotype of a normal cell. (b)
contribute to the development of cancer? Karyotype of a cancer cell showing translocations, deletions, and
aneuploidycharacteristic features of cancer cells.
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 359
16.1
TABLE 16.1 Cancer Probabilities in the United States
Cancer Site Gender Birth to 39
All sites Male 1 in 62
Female 1 in 52
Breast Female 1 in 235
Prostate Male 61 in 10,000
Lung, bronchus Male 1 in 3300
Female 1 in 3180
Colon, rectum Male 1 in 1500
Female 1 in 1900
Source: American Cancer Society
ness to treatments. However, at the molecular level, all cancers
exhibit common characteristics that unite them as a family.
All cancer cells share two fundamental properties: (1) abnor-
mal cell growth and division (cell proliferation), and (2) abnor-
malities in the normal restraints that keep cells from spreading
and invading other parts of the body (metastasis). In normal
cells, these functions are tightly controlled by genes that are
expressed appropriately in time and place. In cancer cells, these
genes are either mutated or are expressed inappropriately.
It is this combination of uncontrolled cell proliferation and
metastatic spread that makes cancer cells dangerous. When a
cell simply loses genetic control over cell growth, it may grow
into a multicellular mass, a benign tumor. Such a tumor can
often be removed by surgery and may cause no serious harm.
However, if cells in the tumor also acquire the ability to break
loose, enter the bloodstream, invade other tissues, and form
secondary tumors (metastases), they become malignant.
Malignant tumors are difficult to treat and may become life
threatening. As we will see later in the chapter, there are mul-
tiple steps and genetic mutations that convert a benign tumor
into a dangerous malignant tumor.
The Clonal Origin of Cancer Cells
Although malignant tumors may contain billions of cells, and
may invade and grow in numerous parts of the body, all can-
cer cells in the primary and secondary tumors are clonal,
meaning that they originated from a common ancestral cell
that accumulated numerous specific mutations. This is an
important concept in understanding the molecular causes of
cancer and has implications for its diagnosis.
Numerous data support the concept of cancer clonality. For
example, reciprocal chromosomal translocations are characteris-
tic of many cancers, including leukemias and lymphomas (two
cancers involving white blood cells). Cancer cells from patients
with Burkitts lymphoma show reciprocal translocations
between chromosome 8 (with translocation breakpoints at or
near the c-myc gene) and chromosomes 2, 14, or 22 (with
translocation breakpoints at or near one of the immunoglobulin
genes). Each Burkitts lymphoma patient exhibits unique break-
Cancer Is a Genetic Disease at the Level of Somatic Cells 359
Age
4059 6079 Birth to Death
1 in 12 1 in 3 1 in 2
1 in 11 1 in 4 1 in 3
1 in 25 1 in 15 1 in 8
1 in 53 1 in 7 1 in 6
1 in 92 1 in 17 1 in 13
1 in 120 1 in 25 1 in 17
1 in 124 1 in 29 1 in 18
1 in 149 1 in 33 1 in 18
points in their c-myc and immunoglobulin gene DNA sequences;
however, all lymphoma cells within that patient contain identical
translocation breakpoints. This demonstrates that all cancer cells
in each case of Burkitts lymphoma arise from a single cell, and
this cell passes on its genetic aberrations to its progeny.
Another demonstration that cancer cells are clonal is their pat-
tern of X-chromosome inactivation. As explained in Chapter 5,
female humans are mosaic, with some cells containing an inacti-
vated paternal X chromosome and other cells containing an inac-
tivated maternal X chromosome. X-chromosome inactivation
occurs early in development and occurs at random. All cancer
cells within a tumor, both primary and metastatic, within one
female individual, contain the same inactivated X chromosome.
This supports the concept that all the cancer cells in that patient
arose from a common ancestral cell.
Cancer As a Multistep Process, Requiring
Multiple Mutations
Although we know that cancer is a genetic disease initiated by
mutations that lead to uncontrolled cell proliferation and
metastasis, a single mutation is not sufficient to transform a
normal cell into a tumor-forming (tumorigenic), malignant
cell. If it were sufficient, then cancer would be far more preva-
lent than it is. In humans, mutations occur spontaneously at a
rate of about 10-6 mutations per gene, per cell division, mainly
due to the intrinsic error rates of DNA replication. As there
are approximately 1016 cell divisions in a human body during
a lifetime, a person might suffer up to 1010 mutations per gene
somewhere in the body, during his or her lifetime. However,
only about one person in three will suffer from cancer.
The phenomenon of age-related cancer is another indication
that cancer develops from the accumulation of several muta-
genic events in a single cell. The incidence of most cancers
rises exponentially with age. If only a single mutation were
sufficient to convert a normal cell to a malignant one, then can-
cer incidence would appear to be independent of age. The age-
related incidence of cancer suggests that up to 10 independent
mutations, occurring randomly and with a low probability, are
necessary before a cell is transformed into a malignant cancer
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 360

360 Chapter 16 Cell-Cycle Regulation and Cancer

cell. Another indication that cancer is a multistep process is the (a) Double minutes (b) Heterogeneous staining region
delay that occurs between exposure to carcinogens (cancer-
causing agents) and the appearance of the cancer. For example,
an incubation period of five to eight years separated exposure
of people to the radiation of the atomic explosions at Hiroshima
and Nagasaki and the onset of leukemias.
The multistep nature of cancer development is supported by
the observation that cancers often develop in progressive steps,
from tumors containing mildly aberrant cells to those that are
increasingly tumorigenic and malignant. This progressive
nature of cancer is illustrated by the development of colon can-
cer, as discussed later in this chapter (see Section 16.6).
Each step in tumorigenesis (the development of a malignant
FIGURE 162 DNA amplifications in neuroblastoma cells. (a) Two
tumor) appears to be the result of one or more genetic alterations cancer genes (MYCN in red and MDM2 in green) are amplified as small
that release the cells progressively from the controls that normal- DNA fragments that remain separate from chromosomal DNA within the
ly operate on proliferation and malignancy. As we will discover nucleus. These units of amplified DNA are known as double minute
in subsequent sections of this chapter, the genes that undergo chromosomes. Normal chromosomes are stained blue. (b) Multiple
mutations leading to cancer (called oncogenes and tumor sup- copies of the MYCN gene are amplified within one large region called a
pressor genes) are those that control DNA damage repair, the cell heterogeneous staining region (green). Single copies of the MYCN gene
cycle, cellcell contact, and programmed cell death. are visible as green dots at the ends of the normal parental chromosomes
We will now investigate each of these fundamental processes, (white arrows). Normal chromosomes are stained red.
the genes that control them and how mutations in these genes
may lead to cancer. CML cells, constantly stimulating these cells to proliferate even in
the absence of external growth signals.
In keeping with the concept of the cancer mutator phenotype, a
16.2 Cancer Cells Contain Genetic number of inherited cancers are caused by defects in genes that
control DNA repair. For example, xeroderma pigmentosum (XP)
Defects Affecting Genomic
is a rare hereditary disorder that is characterized by extreme sen-
Stability and DNA Repair sitivity to ultraviolet light and other carcinogens. Patients with
Cancer cells show higher than normal rates of mutation, chro- XP often develop skin cancer. Cells from patients with XP are
mosomal abnormalities, and genomic instability. In fact, many defective in nucleotide excision repair, with mutations appearing
researchers believe that the fundamental defect in cancer cells is in any one of seven genes whose products are necessary to carry
a derangement of the cells normal ability to repair DNA dam- out DNA repair. XP cells are impaired in their ability to repair
age. This loss of genomic integrity leads to a general increase in DNA lesions such as thymine dimers induced by UV light. The
the mutation rate in every gene, including specific genes that
control aspects of cell proliferation, programmed cell death, and
Normal Translocation
cellcell contact. In turn, the accumulation of mutations in genes chromosome 9 t(9;22)
controlling these processes leads to cancer. The high level of
genomic instability seen in cancer cells is known as the mutator
phenotype. Normal
chromosome 22
Genomic instability in cancer cells manifests itself by the pres-
ence of gross defects such as translocations, aneuploidy, chromo-
some loss, DNA amplification, and chromosome deletions
(Figures 161 and 162). Cancer cells that are grown in cultures
q11.2 (BCR)
in the lab also show a great deal of genomic instability
(BCR) (ABL)
duplicating, losing, and translocating chromosomes or parts of
chromosomes. Often cancer cells show specific chromosomal Philadelphia
defects that are used to diagnose the type and stage of the cancer. chromosome
For example, leukemic white blood cells from patients with q34.1 (C-ABL)
chronic myelogenous leukemia (CML) bear a specific translo-
cation, in which the C-ABL gene on chromosome 9 is translocat-
FIGURE 163 A reciprocal translocation involving the long arms of
ed into the BCR gene on chromosome 22. This translocation
chromosomes 9 and 22 results in the formation of a characteristic
creates a structure known as the Philadelphia chromosome
chromosome, the Philadelphia chromosome, which is associated with
(Figure 163). The BCR-ABL fusion gene codes for a chimeric chronic myelogenous leukemia (CML). The t(9;22) translocation
BCR-ABL protein. The normal ABL protein is a protein kinase results in the fusion of the C-ABL proto-oncogene on chromosome 9
that acts within signal transduction pathways, transferring growth with the BCR gene on chromosome 22. The fusion protein is a
factor signals from the external environment to the nucleus. The powerful hybrid molecule that allows cells to escape control of the
BCR-ABL protein is an abnormal signal transduction molecule in cell cycle, contributing to the development of CML.
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 361

16.3 Cancer Cells Contain Genetic Defects Affecting Cell-Cycle Regulation 361

C,S C

C S C,E C B C,S C,E

C C C C,P C C C C C C C

C
FIGURE 164 Pedigree of a family with HNPCC. Families with HNPCC are defined as those in which at
least three relatives in two generations have been diagnosed with colon cancer, with one relative diagnosed
at less than 50 years of age. Colon cancer: C; stomach cancer: S; endometrial cancer: E; pancreatic cancer:
P; bladder/urinary cancer: B. Blue symbols indicate family members with colon cancer; diagonal stripes
mean that diagnosis is uncertain; orange symbols indicate other tumors. Symbols with slashes indicate
deceased individuals. Reprinted with permission from Aaltonen et al. Clues to the pathogenesis of familial
colorectal cancer. Science 260: 812-816. Copyright 1993 AAAS.

relationship between XP and genes controlling nucleotide exci-

sion repair is also described in Chapter 14.
Another hereditary cancer, hereditary nonpolyposis colorectal
cancer (HNPCC), is also caused by mutations in genes control-
ling DNA repair. HNPCC is an autosomal dominant syndrome,
affecting about one in every 200 people (Figure 164). Patients
affected by HNPCC have an increased risk of developing colon,
ovary, uterine, and kidney cancers. Cells from patients with
HNPCC show higher than normal mutation rates and genomic
instability. At least eight genes are associated with HNPCC, and
four of these genes control aspects of DNA mismatch repair. Inac-
tivation of any of these four genesMSH2, MSH6, MLH1, and
MLH3causes a rapid accumulation of genome-wide mutations
and the subsequent development of colorectal and other cancers.
The observation that hereditary defects in genes controlling
nucleotide excision repair and DNA mismatch repair lead to high
rates of cancer lends support to the idea that the mutator pheno-
type is a significant contributor to the development of cancer.
Now Solve This
Problem 17 on page 374 asks you to consider how the BCR-
ABL hybrid fusion protein, found in CML leukemic white
blood cells, could be used as a target for cancer therapy.
Hint: Most cancer therapies, including radiation and
chemotherapies, aim to kill cells that are constantly
dividing. However, many normal cells in the body also
divide and are killed by cancer therapies, leading to side
effects. The BCR-ABL fusion protein is found only in CML
white blood cells. As you try to answer this problem, you
may wish to learn more about the drug Gleevec (see
http://www.nci.nih.gov/newscenter/qandagleevec).
16.3 Cancer Cells Contain Genetic
Defects Affecting Cell-Cycle
Regulation
One of the fundamental aberrations in all cancer cells is a loss
of control over cell proliferation. Cell proliferation is the
process of cell growth and division that is essential for all
development and tissue repair in multicellular organisms.
Although some cells, such as epidermal cells of the skin or
blood-forming cells in the bone marrow, continue to grow and
divide throughout the organisms lifetime, most cells in adult
multicellular organisms are in a nondividing, quiescent, and
differentiated state. Differentiated cells are those that are spe-
cialized for a specific function, such as photoreceptor cells of
the retina or muscle cells of the heart. The most extreme exam-
ples of nonproliferating cells are nerve cells, which little, if at
all, even to replace damaged tissue. In contrast, many differen-
tiated cells, such as those in the liver and kidney, are able to
grow and divide when stimulated by extracellular signals and
growth factors. In this way, multicellular organisms are able to
replace dead and damaged tissue. However, the growth and dif-
ferentiation of cells must be strictly regulated; otherwise, the
integrity of organs and tissues would be compromised by inap-
propriate types and quantities of cells. Normal regulation over
cell proliferation involves a large number of gene products that
control steps in the cell cycle, programmed cell death, and the
response of cells to external growth signals. In cancer cells,
many of the genes that control these functions are mutated or
aberrantly expressed, leading to uncontrolled cell proliferation.
In this section, we will review steps in the cell cycle, some
of the genes that control the cell cycle, and how these genes,
when mutated, lead to cancer.
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 362

362 Chapter 16 Cell-Cycle Regulation and Cancer

The Cell Cycle and Signal Transduction nucleus even in the absence of external growth signals. In
The cellular events that occur from one cell division to the addition, malignant cells may not respond to external signals
next comprise the cell cycle (Figure 165). The interphase from surrounding cellssignals that would normally inhibit
stage of the cell cycle is the interval between mitotic divi- cell proliferation within a mature tissue.
sions. During this time, the cell grows and replicates its DNA.
During G1, the cell prepares for DNA synthesis by accumu- Cell-Cycle Control and Checkpoints
lating the enzymes and molecules required for DNA replica-
In normal cells, progress through the cell cycle is tightly regu-
tion. G1 is followed by S phase, during which the cells
lated, and the completion of each step is necessary prior to ini-
chromosomal DNA is replicated. During G2, the cell contin-
tiating the next step. There are at least three distinct points in
ues to grow and prepare for division. During M phase, the
the cell cycle at which the cell monitors external signals and
duplicated chromosomes condense, sister chromosomes sepa-
internal equilibrium before proceeding to the next stage of the
rate to opposite poles, and the cell divides in two.
cell cycle. These are the G1/S, the G2/M, and M checkpoints
In early to mid-G1, the cell makes a decision either to enter
(Figure 165). At the G1/S checkpoint, the cell monitors its
the next cell cycle or to withdraw from the cell cycle into qui-
size and determines if its DNA has been damaged. If the cell
escence. Continuously dividing cells do not exit the cell cycle
has not achieved an adequate size, or if the DNA has been
but proceed through G1, S, G2, and M phases until they
damaged, further progress through the cell cycle is halted until
receive signals to stop growing. If the cell stops proliferating,
these conditions are corrected. If cell size and DNA integrity
it enters the G0 phase of the cell cycle. During G0, the cell
are normal, the G1/S checkpoint is traversed, and the cell pro-
remains metabolically active but does not grow or divide.
ceeds to S phase. The second important checkpoint is the
Most differentiated cells in multicellular organisms can
G2/M checkpoint, where physiological conditions in the cell
remain in this G0 phase indefinitely. Some, such as neurons,
are monitored prior to entering mitosis. If DNA replication or
never reenter the cell cycle. In contrast, cancer cells are
repair of any DNA damage has not been completed, the cell
unable to enter G0 and continuously cycle. Their rate of pro-
cycle arrests until these processes are complete. The third
liferation is not necessarily any greater than a normal prolifer-
major checkpoint occurs during mitosis and is called the M
ating cell; however, they are not able to become quiescent at
checkpoint. At this checkpoint, both the successful formation
the appropriate time or place.
of the spindle-fiber system and the attachment of spindle fibers
Cells in G0 can often be stimulated to reenter the cell cycle
to the kinetochores associated with the centromeres are moni-
by external growth signals. These signals are delivered to the
tored. If spindle fibers are not properly formed or attachment is
cell by molecules such as growth factors and hormones that
inadequate, mitosis is arrested.
bind to cell-surface receptors, which then relay the signal
In addition to regulating the cell cycle at checkpoints, the
from the plasma membrane to signal transduction molecules
cell controls progress through the cell cycle through two
located in the cytoplasm. Ultimately, signal transduction initi-
classes of proteins: cyclins and cyclin-dependent kinases
ates a program of gene expression that propels the cell out of
(CDKs). The cell synthesizes and destroys cyclin proteins in
G0 back into the cell cycle. Cancer cells often have defects in
a precise pattern during the cell cycle (Figure 166). When a
signal transduction pathways. Sometimes, abnormal signal
cyclin is present, it binds to a specific CDK, triggering activ-
transduction molecules send continuous growth signals to the
ity of the CDK/cyclin complex. The CDK/cyclin complex-
es then selectively phosphorylate and activate other
proteins that in turn bring about the changes necessary to
G1/S checkpoint
advance the cell through the cell cycle. For example, in
Cell monitors size and G1 phase, CDK4/cyclin D complexes activate proteins
DNA integrity that stimulate transcription of genes whose products (such
G1 S as DNA polymerase d and DNA ligase) are required for
Decision to reenter DNA replication during S phase. Another CDK/cyclin
cell cycle
Pro complex, CDK1/cyclin B, phosphorylates a number of
iescence

proteins that bring about the events of early mitosis, such

li f e

G0 rat
ion as nuclear membrane breakdown, chromosome condensa-
tion and cytoskeletal reorganization. Mitosis can only be
Qu

completed, however, when cyclin B is degraded and the

Decision to exit G1 protein phosphorylations characteristic of M phase are
cell cycle G2 G2/M checkpoint
M reversed. Although a large number of different protein
Cell monitors DNA kinases exist in cells, only a few are involved in cell-cycle
M checkpoint synthesis and damage
Cell monitors regulation.
spindle formation Both cell-cycle checkpoints and cell-cycle control mole-
and attachment to cules are genetically regulated. In general, the cell cycle is
kinetochores
regulated by an interplay of genes whose products either pro-
FIGURE 165 Checkpoints and proliferation decision points mote or suppress cell division. Mutation or mis-expression of
monitor the progress of the cell through the cell cycle. any of the genes controlling the cell cycle contributes to the
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 363

16.3 Cancer Cells Contain Genetic Defects Affecting Cell-Cycle Regulation 363

FIGURE 166 Relative expression times and amounts of

cyclins during the cell cycle. Cyclin D1 accumulates early
in G1 and is expressed at a constant level through most of
Relative amounts of cyclins

the cycle. Cyclin E accumulates in G1, reaches a peak, and

D2 A declines by mid-S phase. Cyclin D2 begins accumulating in
B the last half of G1, reaches a peak just after the beginning
D1
of S, and then declines by early G2. Cyclin A appears in
late G1, accumulates through S phase, peaks at the G2/M
transition, and is rapidly degraded. Cyclin B peaks at the
E G2/M transition and declines rapidly in M phase.

G1 S G2 M
Phases of the cell cycle

development of cancer in several ways. For example, if genes
that control the G1/S or G2/M checkpoints are mutated, the
cell may continue to cycle before repairing DNA damage.
This may lead to the accumulation of mutations in genes
whose product control cell proliferation or metastasis. Simi-
larly, if genes that control progress through the cell cycle,
such as those that encode the cyclins, are expressed inappro-
priately, the cell may cycle continuously and may be unable to
exit the cell cycle into G0.
As already described, if DNA replication, repair, or chromo-
some assembly is aberrant, normal cells halt their progress
through the cell cycle until the condition is corrected. This
reduces the number of mutations and chromosomal abnormal-
ities that accumulate in normal proliferating cells. However, if
DNA or chromosomal damage is so severe that repair is
impossible, the cell may initiate a second line of defensea
process called apoptosis, or programmed cell death. Apop-
tosis is a genetically controlled process whereby the cell
commits suicide. Apoptosis is also initiated during normal
multicellular development in order to eliminate certain cells
that do not contribute to the final adult organism. The steps in
apoptosis are the same for damaged cells and for cells elimi-
nated during development: Nuclear DNA becomes fragment-
ed, internal cellular structures are disrupted, and the cell
dissolves into small spherical structures known as apoptotic
bodies (Figure 167). In the final step, the apoptotic bodies
are engulfed by the immune systems phagocytic cells. A
series of proteases called caspases are responsible for initiat-
ing apoptosis and for digesting intracellular components.
Apoptosis is genetically controlled in that regulation of spe-
cific gene products such as the Bcl2 and BAX proteins can
trigger or prevent apoptosis. By removing damaged cells,

(a) (b)

Bcl2 Bcl2-BAX BAX

Homodimer Heterodimer Homodimer

Blocks Inactive Promotes

Apoptosis Complex Apoptosis

FIGURE 167 (a) A normal white blood cell (bottom) and a white blood cell undergoing apoptosis (top).
Apoptotic bodies appear as grape-like clusters on the cell surface. (b) The relative concentrations of the
Bcl2 and BAX proteins regulate apoptosis. A normal cell contains a balance of Bcl2 and BAX, which
form inactive heterodimers. A relative excess of Bcl2 results in Bcl2 homodimers, which prevent
apoptosis. Cancer cells with Bcl2 overexpression are resistant to chemotherapies and radiation therapies.
A relative excess of BAX results in BAX homodimers, which induce apoptosis. In normal cells, activated
p53 protein induces transcription of BAX and inhibits transcription of Bcl2, leading to cell death. In many
cancer cells, p53 is defective, preventing the apoptotic pathway from removing the cancer cells.
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 364

364 Chapter 16 Cell-Cycle Regulation and Cancer

programmed cell death reduces the number of mutations that
are propagated to the next generation, including those in can-
cer-causing genes. The same genes that regulate cell-cycle
checkpoints can trigger apoptosis. As we will see, these genes
are mutated in a many cancers.
16.4 Many Cancer-Causing Genes
Disrupt Control of the Cell Cycle
Two general categories of genes are mutated or mis-expressed
in cancer cellsthe proto-oncogenes and the tumor suppres-
sor genes (Table 16.2). Proto-oncogenes are genes whose
products are important for normal cell functions and promote
cell growth and division. They do this by encoding transcrip-
tion factors that stimulate expression of other genes, signal
tranduction molecules that stimulate cell division, or cell-
cycle regulators that move the cell through the cell cycle.
When normal cells become quiescent and cease division, they
repress the expression or activity of most proto-oncogene
products. In cancer cells, one or more proto-oncogenes are
altered in such a way that their activities cannot be controlled
in a normal fashion. This is sometimes due to a mutation in
the proto-oncogene resulting in a protein product that acts
abnormally. In other cases, proto-oncogenes are overex-
pressed or cannot be transcriptionally repressed at the correct
time. In these cases, the proto-oncogene is continually in an
on state, which may constantly stimulate the cell to divide.
When a proto-oncogene is mutated or aberrantly expressed
and contributes to the development of cancer, it is known as
an oncogene (cancer-causing gene). Oncogenes are those that
have experienced a gain-of-function alteration. As a result,
only one allele of a proto-oncogene needs to be mutated or
mis-expressed in order to trigger uncontrolled growth. Hence,
oncogenes confer a dominant cancer phenotype.
Tumor suppressor genes are those whose products normal-
ly regulate cell-cycle checkpoints and initiate the process of
apoptosis. In normal cells, proteins encoded by tumor sup-
pressor genes halt progress through the cell cycle in response
to DNA damage or growth-suppression signals from the
extracellular environment. When tumor suppressor genes are
mutated or inactivated, cells are unable to respond normally
to cell-cycle checkpoints, or are unable to undergo pro-
grammed cell death if DNA damage is extensive. This leads to
a further increase in mutations and to the inability of the cell
to leave the cell cycle when it should become quiescent.

TABLE 16.2 Some Proto-oncogenes and Tumor Suppressor Genes

Proto-oncogene Normal Function Alteration in Cancer Associated Cancers

Ha-ras Signal transduction molecule, binds Point mutations Colorectal, bladder, many
GTP/GDP types

c-erbB Transmembrane growth factor receptor Gene amplification, Glioblastomas, breast cancer,
point mutations cervix

c-myc Transcription factor, regulates cell cycle, Translocation, amplification, Lymphomas, leukemias, lung
differentiation, apoptosis point mutations cancer, many types

c-kit Tyrosine kinase, signal transduction Mutation Sarcomas

RARa Hormone-dependent transcription Chromosomal translocations with Acute promyelocytic

factor, differentiation PML gene, fusion product leukemia

E6 Human papillomavirus encoded HPV infection Cervical cancer

oncogene, inactivates p53

Cyclins Bind to CDKs, regulate cell cycle Gene amplification, overexpression Lung, esophagus, many types

CDK2, 4 Cyclin-dependent kinases, regulate Overexpression, mutation Bladder, breast, many types
cell-cycle phases

Tumor Suppressor Normal Function Alteration in Cancer Associated Cancers

p53 Cell-cycle checkpoints, apoptosis Mutation, inactivation by viral Brain, lung, colorectal,
oncogene products breast, many types

RB1 Cell-cycle checkpoints, binds E2F Mutation, deletion, inactivation by Retinoblastoma,

viral oncogene products osteosarcoma, many types

APC Cellcell interaction Mutation Colorectal cancers, brain,

thyroid

Bcl2 Apoptosis regulation Overexpression blocks apoptosis Lymphomas, leukemias

BRCA2 DNA repair Point mutations Breast, ovarian, prostate

cancers
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 365

16.4 Many Cancer-Causing Genes Disrupt Control of the Cell Cycle 365

When both alleles of a tumor suppressor gene are inactivated, 1. Growth factor binds to
Growth
and other changes in the cell keep it growing and dividing, factor cell surface receptor
cells may become tumorigenic.
The following are examples of proto-oncogenes and tumor Receptor PLASMA MEMBRANE
suppressor genes that contribute to cancer when mutated.
There are more than 200 oncogenes and tumor suppressor
genes now known, and more will likely be discovered as can-
AU: already appeared as K-Term. Set normal font?

cer research continues. GDPGTP

exchange factor
The ras Proto-oncogenes
Ras Ras
Some of the most frequently mutated genes in human tumors 2. Ras transiently exchanges
are those of the ras gene family. These genes are mutated in GDP GTP GTP for GDP
more than 40 percent of human tumors. The ras gene family
Inactive Active
encodes signal transduction molecules that are associated
3. Ras sends signals to
with the cell membrane and regulate cell growth and division. cascades of activated
Ras proteins normally transmit signals from the cell mem- proteins
Raf
brane to the nucleus, stimulating the cell to divide in response
to external growth factors. Ras proteins alternate between an
inactive (switched off) and an active (switched on) state by
binding either guanosine diphosphate (GDP) or guanosine Mek CYTOPLASM
triphosphate (GTP). When a cell encounters a growth factor
(such as platelet-derived growth factor or epidermal growth
factor), growth factor receptors on the cell membrane bind to
the growth factor, resulting in autophosphorylation of the cyto- Map Kinase
plasmic portion of the growth factor receptor. This causes 4. Signal transduction proteins
recruitment of proteins known as nucleotide exchange factors activate transcription factors
to the plasma membrane. These nucleotide exchange factors
Transcription
cause Ras to release GDP and bind GTP, thereby activating Factors
Ras. The active, GTP-bound form of Ras then sends its signals
through cascades of protein phosphorylations in the cytoplasm NUCLEUS 5. Activation or repression
of gene transcription
(Figure 168). The end-point of these cascades is activation of
nuclear transcription factors that stimulate expression of genes
whose products drive the cell from quiescence into the cell
cycle. Once Ras has sent its signals to the nucleus, it FIGURE 168 A signal transduction pathway mediated by Ras.
hydrolyzes GTP to GDP and becomes inactive. Mutations that
convert the proto-oncogene ras to an oncogene prevent the Ras
from increases in protein phosphorylation, acetylation, and
protein from hydrolyzing GTP to GDP and hence freeze the
p53 protein stability.
Ras protein into its on conformation, constantly stimulating
The p53 protein initiates two different responses to DNA
the cell to divide.
damage: arrest of the cell cycle followed by DNA repair, or
apoptosis and cell death if DNA cannot be repaired. Each of
The p53 Tumor Suppressor Gene these responses is accomplished by p53 acting as a transcrip-
The most frequently mutated gene in human cancersoccurring tion factor that stimulates or represses the expression of genes
in more than 50 percent of all cancersis the p53 gene. This involved in each of these responses.
gene encodes a nuclear protein that acts as a transcription fac- In normal cells, p53 can arrest the cell cycle at several phases.
tor that represses or stimulates transcription of more than 50 To arrest the cell cycle at the G1/S checkpoint, activated p53
different genes. protein stimulates transcription of a gene encoding the p21 pro-
Normally, the p53 protein is continuously synthesized but is tein. The p21 protein inhibits the CDK4/cyclin D1 complex,
rapidly degraded and therefore is present in cells at low levels. hence preventing the cell from moving from the G1 phase into S
In addition, the p53 protein is normally bound to another pro- phase. Activated p53 protein also regulates expression of genes
tein called Mdm2, which prevents the phosphorylations and that retard the progress of DNA replication, thus allowing time
acetylations that convert the p53 protein from an inactive to for DNA damage to be repaired during S phase. By regulating
an active form. Several types of events cause a rapid increase expression of other genes, activated p53 can block cells at the
in the nuclear levels of activated p53 protein. These include G2/M checkpoint, if DNA damage occurs during S phase.
chemical damage to DNA, double-stranded breaks in DNA Activated p53 can also instruct a damaged cell to commit
induced by ionizing radiation, or the presence of DNA-repair suicide by apoptosis. It does so by activating the transcription
intermediates generated by exposure of cells to ultraviolet of the Bax gene and repressing transcription of the Bcl2 gene.
light. Increases in the levels of activated p53 protein result In normal cells, the BAX protein is present in a heterodimer
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 366

366 Chapter 16 Cell-Cycle Regulation and Cancer

with the Bcl2 protein and the cell remains viable. When the
levels of BAX protein increase in response to p53 stimulation
of Bax gene transcription, BAX homodimers are formed and
these homodimers activate the cellular changes that lead to
cellular self-destruction (Figure 167). In cancer cells that
lack functional p53, BAX protein levels do not increase in
response to cellular damage and apoptosis may not occur.
Hence, cells lacking functional p53 are unable to arrest at
Web Tutorial 16.1
oping retinoblastomas as well as an increased chance of devel-
oping other cancers. All somatic cells of patients with heredi-
tary retinoblastoma contain one mutated allele of the RB1 gene.
However, it is only when the second normal allele of the RB1
gene is lost or mutated in certain retinal cells that retinoblas-
toma develops. In individuals that do not have this hereditary
condition, retinoblastoma is extremely rare, as it requires at
least two separate somatic mutations in a retinal cell in order to

cell-cycle checkpoints or to enter apoptosis in response to inactivate both copies of the RB1 gene (Figure 169).
DNA damage. As a result, they move unchecked through the The retinoblastoma protein (pRB) is a tumor suppressor
Retinoblastoma

cell cycle, regardless of the condition of the cells DNA. Cells
lacking p53 have high mutation rates and accumulate the
types of mutations that lead to cancer. Because of the impor-
tance of the p53 gene to genomic integrity, it is often referred
to as the guardian of the genome.
The RB1 Tumor Suppressor Gene
The loss or mutation of the RB1 (retinoblastoma 1) tumor
suppressor gene contributes to the development of many can-
cers, including those of the breast, bone, lung, and bladder. The
RB1 gene was originally identified as a result of studies on
retinoblastoma, an inherited disorder in which tumors develop
in the eyes of young children. Retinoblastoma occurs with a
frequency of about 1 in 14,000 to 20,000 individuals. In the
familial form of the disease, individuals inherit one mutated
allele of the RB1 gene and have an 85 percent chance of devel-
protein that controls the G1/S cell-cycle checkpoint. The pRB
protein is found in the nuclei of all cell types at all stages of the
cell cycle. However, its activity varies throughout the cell
cycle, depending on its phosphorylation state. When cells are
in the G0 phase of the cell cycle, the pRB protein is nonphos-
phorylated and binds to transcription factors such as E2F, inac-
tivating them (Figure 1610). When the cell is stimulated by
growth factors, it enters G1 and approaches S phase. Through-
out the G1 phase, the pRB protein becomes phosphorylated by
the CDK4/cyclin D1 complex. Phosphorylated pRB is inactive
and releases its bound regulatory proteins. When E2F and
other regulators are released by pRB, they are free to induce
the expression of over 30 genes whose products are required
for the transition from G1 into S phase. After cells traverse S,
G2, and M phases, pRB reverts to a nonphosphorylated state,
binds to regulatory proteins such as E2F, and keeps them

(a) Familial retinoblastoma (b) Sporadic retinoblastoma FIGURE 169 (a) In

familial retinoblastoma, one
mutation (designated as
Cell with inherited Normal RB1) is inherited and
RB1/ /
RB1 mutation cell present in all cells. A second
mutation at the
No First spontaneous RBI
retinoblastoma locus in any
mutation mutation
retinal cell contributes to
uncontrolled cell growth and
RB1/ RB1/ / RB1/ tumor formation. (b) In
sporadic retinoblastoma,
two independent mutations
in both alleles of the
retinoblastoma gene within
Spontaneous RB1 a single cell are acquired
RB1/ RB1/ mutation / / RB1/ RB1/ sequentially, also leading to
tumor formation.
RB
1/

Controlled growth, Controlled growth, Second spontaneous

RB
1

no tumor formation no tumor formation RB1 mutation

RB
1
/R

1
RB
B1

/
RB1 RB1
1/
RB

RB
RB

RB1
1

1/

/RB
/R

1
RB
B1

1
RB
1/
RB
1

Uncontrolled growth,
/RB

/RB
1

tumor formation
RB1

RB1

Uncontrolled growth,
tumor formation
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 367
E2F
Inactive
transcription factor
G1 pRB
CDK4/cyclin D1 PO4
complex
phosphorylates pRB PO4
pRB
S erty of some normal cell types. For example, implantation of
E2F Active transcription factor
Gene expression: cell
progresses through cell cycle
Target gene
FIGURE 1610 In the nucleus during G1, pRB interacts with and
inactivates transcription factor E2F. As the cell moves from G1 to S
phase, a CDK4/cyclinD1 complex forms and adds phosphate groups
to pRB. As pRB becomes phosphorylated, E2F is released and
becomes transcriptionally active, allowing the cell to pass through S
phase. Phosphorylation of pRB is transitory; as CDK/cyclin
complexes are degraded and the cell moves through the cell cycle to
early G1, pRB phosphorylation declines.
sequestered until required for the next cell cycle. In normal
quiescent cells, the pRB protein is active and prevents passage
into S phase. In many cancer cells, including retinoblastoma
cells, both copies of the RB1 gene are defective, inactive, or
absent, and progression through the cell cycle is not regulated.
Now Solve This
Problem 23 on page 374 asks you to explain how in the
inherited LiFraumeni syndrome, mutations in one
allele of the p53 gene can give rise to a wide variety of
different cancers.
Hint: To answer this question, you might review the
cellular functions regulated by the normal p53 tumor
suppressor gene. Consider how each of these func-
tions could be affected if the p53 gene product is
either absent or defective. To understand how muta-
tions in one allele of a tumor suppressor gene result in
tumors that contain aberrations in both alleles, read
about loss of heterozygosity in Section 16.6.
16.5 Cancer Is a Genetic Disorder
Affecting Cell Adhesion
As discussed at the beginning of this chapter, uncontrolled
growth alone is insufficient to create a malignant and life-
threatening cancer. Cancer cells must also acquire the features
of metastasis, which include the ability to disengage from the
original tumor site, to enter the blood or lymphatic system, to
16.6 Predisposition to Some Cancers Can Be Inherited 367
invade surrounding tissues, and to develop into secondary
tumors. In order to leave the site of the primary tumor and
invade other tissues, tumor cells must dissociate from other
cells and digest components of the extracellular matrix and
basal lamina, which normally contain and separate tissues.
The extracellular matrix and basal lamina are composed of pro-
teins and carbohydrates. They form the scaffold for tissue
growth and normally inhibit the migration of cells.
The ability to invade the extracellular matrix is also a prop-
the embryo in the uterine wall during pregnancy requires cell
migration across the extracellular matrix. In addition, white
blood cells reach the site of infection by penetrating capillary
walls. The mechanisms of invasion are probably similar in
these normal cells and in cancer cells. The difference is that,
in normal cells, the invasive ability is tightly regulated,
whereas in tumor cells, this regulation has been lost.
Although less is known about the genes that control metasta-
sis than about those controlling the cell cycle, it is likely that
metastasis is controlled by a large number of genes, including
those that encode cell-adhesion molecules, cytoskeleton regu-
lators, and proteolytic enzymes. For example, epithelial tumors
have a lower than normal level of the E-cadherin glycoprotein,
which is responsible for cellcell adhesion in normal tissues.
Also, proteolytic enzymes such as metalloproteinases are pre-
sent at higher than normal levels in highly malignant tumors
and are not susceptible to the normal controls conferred by reg-
ulatory molecules such as tissue inhibitors of metallopro-
teinases (TIMPs). It has been shown that the level of
aggressiveness of a tumor correlates positively with the levels
of proteolytic enzymes expressed by the tumor. Hence, inap-
propriately expressed cell adhesion and proteinase enzymes
may assist malignant tumor cells by loosening the normal con-
straints on cell location and creating holes through which the
tumor cells can pass into and out of the circulatory system.
Now Solve This
Problem 25 on page 374 describes the isolation of a
gene that is mutated in metastatic tumors. The gene
appears to be a member of the serine protease family.
You are asked to conjecture how this gene might con-
tribute to the development of highly invasive cancers.
Hint: As you learned in this section, metastatic cancer cells
have the ability to escape their adhesions to other cells,
travel through the circulatory system, and invade distant
tissues. You might think about what types of metastatic
processes are facilitated by having a protease molecule
that lacks normal regulation. Also consider what types of
mutations in this gene could lead to increased metastasis.
16.6 Predisposition to Some Cancers
Can Be Inherited
Although the vast majority of human cancers are sporadic, a
small fraction (12 percent) have an hereditary or familial
component. At present, about 50 forms of hereditary cancer
are known (Table 16.3).
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 368

368 Chapter 16 Cell-Cycle Regulation and Cancer

The development of hereditary colon cancer illustrates

TABLE 16.3 Inherited Predispositions to Cancer
how inherited mutations in one allele of a gene con-
Tumor Predisposition Syndromes Chromosome tribute only one step in the multistep pathway leading to
malignancy.
Early-onset familial breast cancer 17q
About 1 percent of colon cancer cases result from a
Familial adenomatous polyposis 5q genetic predisposition to cancer known as familial ade-
Familial melanoma 9p
nomatous polyposis (FAP). In FAP, individuals inherit
one mutant copy of the APC (adenomatous polyposis)
Gorlin syndrome 9q gene located on the long arm of chromosome 5. Muta-
Hereditary nonpolyposis colon cancer 2p tions include deletions, frameshift, and point mutations.
The normal function of the APC gene product is to act as
Li-Fraumeni syndrome 17p
a tumor suppressor controlling cellcell contact and
Multiple endocrine neoplasia, type 1 11q growth inhibition by interacting with the b -catenin pro-
tein. The presence of a heterozygous APC mutation caus-
Multiple endocrine neoplasia, type 2 22q
es the epithelial cells of the colon to partially escape
Neurofibromatosis, type 1 17q cell-cycle control, and the cells divide to form small clus-
Neurofibromatosis, type 2 22q ters of cells called polyps or adenomas. People who are
heterozygous for this condition develop hundreds to thou-
Retinoblastoma 13q sands of colon and rectal polyps early in life. Although it
Von Hippel-Lindau syndrome 3p is not necessary for the second allele of the APC gene to
be mutated in polyps at this stage, in the majority of
Wilms tumor 11p
cases, the second APC allele becomes mutant in a later
stage of cancer development. The relative order of muta-
Most inherited cancer-susceptibility genes, though trans- tions in the development of FAP is shown in Figure
mitted in a Mendelian dominant fashion, are not sufficient 1611.
in themselves to trigger development of a cancer. At least The second mutation in polyp cells that contain an APC
one other somatic mutation in the other copy of the gene gene mutation occurs in the ras proto-oncogene. The com-
must occur in order to drive a cell toward tumorigenesis. In bined APC and ras gene mutations bring about the develop-
addition, mutations in other genes are usually necessary to ment of intermediate adenomas. Cells within these adenomas
fully express the cancer phenotype. As mentioned above, have defects in normal cell differentiation and will grow in
inherited mutations in the RB1 gene predispose individuals culture, in the absence of contact with other cells and hence
to developing various cancers. Although the normal somat- are described as transformed. The third step toward malig-
ic cells of these patients are heterozygous for the RB1 nancy requires loss of function of both alleles of the DCC
mutation, cells within their tumors contain mutations in (deleted in colon cancer) gene. The DCC gene product is
both copies of the gene. The phenomenon whereby the sec- thought to be involved with cell adhesion and differentiation.
ond, wild-type, allele is mutated in a tumor is known as Mutations in both DCC alleles result in the formation of late-
loss of heterozygosity. Although loss of heterozygosity is stage adenomas with a number of finger-like outgrowths
an essential first step in expression of these inherited can- (villi). When late adenomas progress to cancerous adenomas,
cers, further mutations in other proto-oncogenes and tumor they usually suffer loss of functional p53 genes. The final
suppressor genes are necessary for the passage of tumor steps toward malignancy involve mutations in an unknown
cells to full malignancy. number of genes associated with metastasis.

Chromosome 5q 12p 18q 17p

Alteration Mutation Mutation Loss Loss
Gene APC ras DCC p53

Normal colon Proliferating Benign Intermediate Late adenoma Cancerous Colon

epithelium epithelium adenoma adenoma with villi adenoma cancer

FIGURE 1611 A model for the multistep development of colon cancer. The first step is the loss or inactivation of one allele of the APC
gene on chromosome 5. In FAP cases, one mutant APC allele is inherited. Subsequent mutations involving genes on chromosomes 12, 17,
and 18 in cells of benign adenomas can lead to a malignant transformation that results in colon cancer. Although the mutations on
chromosomes 12, 17, and 18 usually occur at a later stage than those involving chromosome 5, the sum of changes is more important than the
order in which they occur.
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 369
16.7
Now Solve This
Problem 7 on page 372 asks you to explain why some
tobacco smokers and some people with inherited muta-
tions in cancer-related genes never develop cancer.
Hint: When considering the reasons why cancer is not
an inevitable consequence of genetic or environmen-
tal conditions, you might think about the multistep
nature of cancer, the random nature of the sponta-
neous mutations, and the types of genetically con-
trolled functions that are abnormal in cancer cells. It
might also be useful to consider how each individuals
genetic background may affect mutation rates and
differences in DNA repair functions.
16.7 Viruses Contribute to Cancer in
Both Humans and Animals
Viruses that cause cancer in animals have played a significant
role in the search for knowledge about the genetics of human
cancer. Most cancer-causing animal viruses are RNA viruses
known as retroviruses. Because they transform cells into
cancer cells, they are known as acute transforming retroviruses.
The first of these acute transforming retroviruses was dis-
covered in 1910 by Francis Peyton Rous. Rous was studying
sarcomas (solid tumors of muscle, bone, or fat) in chickens,
and he observed that extracts from these tumors caused the
formation of new sarcomas when they were injected into
tumor-free chickens. Several decades later, the agent within
the extract that caused the sarcomas was identified as a retro-
virus and was named the Rous sarcoma virus (RSV).
To understand how retroviruses cause cancer in animals, it is
necessary to know how these viruses replicate in cells. When a
retrovirus infects a cell, its RNA genome is copied into DNA by
Nonacute retrovirus
Genome of virus that can infect
R U5 gag pol env U3 R
but not transform a cell
Cellular proto-oncogene in infected cell
c-onc
Transfer of proto-oncogene from
Acute transforming retrovirus host cell to viral genome
R U5 gag pol env v-onc U3 R
FIGURE 1612 The genome of a typical retrovirus is shown at the
top of the diagram. The genome contains repeats at the termini (R),
the U5 and U3 regions that contain promoter and enhancer
elements, and the three major genes that encode viral structural
proteins (gag and env) and the viral reverse transcriptase (pol).
RNA transcripts of the entire viral genome comprise the new viral
genomes. If the retrovirus acquires all or part of a host cell proto-
oncogene (c-onc), this gene (now known as a v-onc) is expressed
along with the viral genes, leading to overexpression or
inappropriate expression of the v-onc gene. The v-onc gene may
also acquire mutations that enhance its transforming ability.
Viruses Contribute to Cancer in Both Humans and Animals 369
the reverse transcriptase enzyme, which is brought into the
cell with the infecting virus. The DNA copy then enters the
nucleus of the infected cell, where it integrates at random into
the host cells genome. The integrated DNA copy of the retrovi-
ral RNA is called a provirus. The proviral DNA contains pow-
erful enhancer and promoter elements in its U5 and U3
sequences at the ends of the provirus (Figure 1612). The U5
promoter uses the host cells transcription proteins, directing
transcription of the viral genes (gag, pol, and env). The products
of these genes are the proteins and RNA genomes that make up
the new retroviral particles. Because the provirus is integrated
into the host genome, it is replicated along with the hosts DNA
during the cells normal cell cycle. A retrovirus may not kill a
cell, but it may continue to use the cell as a factory to replicate
more viruses that will then infect surrounding cells.
A retrovirus may cause cancer in two different ways. First,
the proviral DNA may integrate by chance near one of the
cells normal proto-oncogenes. The strong promoters and
enhancers in the provirus then stimulate high levels or inap-
propriate timing of transcription of the proto-oncogene, lead-
ing to stimulation of host cell proliferation. Second, a
retrovirus may pick up a copy of a host proto-oncogene and
integrate it into its genome (Figure 1612). The new viral
oncogene may be mutated during the process of transfer into
the virus, or it may be expressed at abnormal levels because it
is now under control of viral promoters. Retroviruses that
carry these viral oncogenes can infect and transform normal
cells into tumor cells. In the case of RSV, the oncogene that
was captured from the chicken genome was the c-src gene.
Through the study of many acute transforming viruses of ani-
mals, scientists have identified dozens of proto-oncogenes.
So far, no acute transforming retroviruses have been identi-
fied in humans. However, RNA and DNA viruses contribute
to the development of human cancers in a variety of ways. It
is thought that, worldwide, about 15 percent of cancers are
associated with viruses, making virus infection the second
greatest risk factor for cancer, next to tobacco smoking.
The most significant contributors to virus-
induced cancers are the papillomaviruses (HPV
16 and 18), human T-cell leukemia virus
(HTLV-1), hepatitis B virus, and Epstein-Barr
virus (Table 16.4). Like other risk factors for can-
cer, including hereditary predisposition to certain
cancers, virus infection alone is not sufficient to
trigger human cancers. Other factors, including
DNA damage or the accumulation of mutations
in one or more of a cells oncogenes and tumor
suppressor genes, are required to move a cell
down the multistep pathway to cancer.
Because viruses are comprised solely of a nucleic acid
genome surrounded by a proteinaceous coat, viruses must
utilize the host cells biosynthetic machinery in order to
reproduce themselves. As most viruses need the host cell to
be in an actively growing state in order to access the hosts
DNA synthetic enzymes, these viruses often contain genes
encoding products that stimulate the cell cycle. If the virus
does not kill the host cell, the potential exists for a loss of cell
cycle control and the beginning of tumorigenesis.
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 370
370 Chapter 16 Cell-Cycle Regulation and Cancer
TABLE 16.4 Human Viruses Associated with Cancer
Virus Cancer
Human papillomavirus 16, 18 Cervical cancer
Hepatitis B virus Liver cancer
Epstein-Barr virus Burkitts lymphoma, naso-pharyngeal
cancer
Human herpesvirus 8 AIDS-related Kaposis sarcoma
Human T-cell leukemia virus Adult T-cell leukemia
16.8 Environmental Agents Contribute
to Human Cancers
Any substance or event that damages DNA has the potential to be
carcinogenic. Unrepaired or inaccurately repaired DNA intro-
duces mutations which, if they occur in proto-oncogenes or
tumor suppressor genes, can lead to abnormal regulation of the
cell cycle or disruption of controls over cell contact and invasion.
Our environment, both natural and man-made, contains
abundant carcinogens. These include chemicals, radiation,
some viruses, and chronic infections. Perhaps the most signif-
icant carcinogen in our environment is tobacco smoke. Epi-
demiologists estimate that about 30 percent of human cancer
deaths are associated with cigarette smoking. Tobacco smoke
contains a number of cancer-causing chemicals, some of
which preferentially mutate proto-oncogenes such as ras and
tumor suppressor genes such as p53.
Diet is often implicated in the development of cancer. Con-
sumption of red meat and animal fat is associated with some
cancers, such as colon, prostate, or breast cancer. The mecha-
nisms by which these substances may contribute to carcino-
genesis are not clear but may involve stimulation of cell
division through hormones or creation of carcinogenic chem-
icals during cooking. Alcohol may cause inflammation of the
liver and contribute to liver cancer.
Although most people perceive the man-made, industrial
environment to be a highly significant contributor to cancer, it
may account for only a small percentage of total cancers, and
often in only specialized situations. Some of the most muta-
genic agents, and hence potentially the most carcinogenic, are
natural substances and natural processes. For example,
aflatoxin, a component of a mold that grows on peanuts and
corn, is one of the most carcinogenic chemicals known. Most
chemical carcinogens, such as nitrosamines, are components
of synthetic substances; however, many are naturally occur-
ring. For example, natural pesticides and antibiotics found in
plants may be carcinogenic, and the human body itself creates
alkylating agents in the acidic environment of the gut. Never-
theless, these observations do not diminish the serious cancer
risks to specific populations who are exposed to man-made
carcinogens, such as synthetic pesticides or asbestos.
DNA lesions brought about by natural radiation (X-rays,
ultraviolet light), natural dietary substances, and the external
AU: missing in K-Term list. Add it?
Oncogenes Mechanism
E6, E7 Inhibit p53 and pRB tumor
suppressors
HBx Signal transduction, stimulates
cell cycle
Unknown Unknown
Several possible Unknown
pX Stimulates cell cycle
environment contribute the majority of environmentally
caused mutations that lead to cancer. Normal metabolism cre-
ates oxidative end products that can damage DNA, proteins,
and lipids. It is estimated that the human body suffers about
10,000 damaging DNA lesions per day due to the actions of
oxygen free radicals. DNA repair enzymes deal successfully
with most of this damage; however, some damage may lead to
permanent mutations. The process of DNA replication itself is
mutagenic. Hence, substances like growth factors or hor-
mones that simulate cell division are ultimately mutagenic
and perhaps carcinogenic. Chronic inflammation due to infec-
tion also stimulates tissue repair and cell division, resulting in
DNA lesions accumulating during replication. These muta-
tions may persist, particularly if cell-cycle checkpoints are
compromised due to mutations or inactivation of tumor sup-
pressor genes such as p53 or RB1.
Both ultraviolet (UV) light and ionizing radiation (such as
X-rays and gamma rays) induce DNA damage. UV damage in
sunlight is well accepted as an inducer of skin cancers. Ioniz-
ing radiation has clearly demonstrated itself as a carcinogen
in studies of populations exposed to neutron and gamma radi-
ation from atomic blasts such as those in Hiroshima and
Nagasaki. Another significant environmental component,
radon gas, may be responsible for up to 50 percent of the ion-
izing radiation exposure for the U.S. population and could
contribute to lung cancers in some populations.
Now Solve This
Problem 16 on page 374 asks you to provide your own
estimate of what percentage of money spent on cancer
research should be devoted to research and education
on preventing cancer and what percentage should be
devoted to research on cancer cures.
Hint: In answering this question, think about the rela-
tive rates of environmentally induced and sponta-
neous cancers. Second, consider the proportion of
environmentally induced cancers that are due to
lifestyle choices. (An interesting source of information
on this topic can be found in Ames, B. N. et al. 1995.
The causes and prevention of cancer. Proc. Natl. Acad.
Sci. USA 92: 52585265.)
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 371
GENETICS, TECHNOLOGY,
Breast Cancer: The Double-Edged Sword of Genetic Testing
These are exhilarat-
ing times for genet-
ics and biotechnology. Close on the heels
of the completion of the Human Genome
Project has come a rush of optimism
about future applications of genetics. Sci-
entists and the media predict that gene
technologies will soon diagnose and cure
diseases as diverse as diabetes, asthma,
heart disease, and Parkinson disease.
The prospect of using genetics to pre-
vent and cure a whole range of diseases is
exciting. However, in our enthusiasm, we
often forget that these new technologies
have significant limitations and profound
ethical concerns. The story of genetic test-
ing for breast cancer illustrates how we
must temper our high expectations with
respect for uncertainty.
Breast cancer is the most common can-
cer among women and the second leading
cause of all cancer deaths (after lung can-
cer). Each year, more than 190,000 new
cases are diagnosed in the United States.
Breast cancer is not limited to women;
about 1400 men are also diagnosed with
the disease each year. A womans lifetime
risk of developing breast cancer is about
12 percent, and the risk increases with
age.
Approximately 5 to 10 percent of breast
cancers are familial, defined by the
appearance of several cases of breast or
ovarian cancer among near blood relatives
and the early onset of these diseases. In
1994, two genes were identified that show
linkage to familial breast cancers. Germ-
line mutations in these genes (BRCA1 and
BRCA2) are associated with the majority
of familial breast cancers. The molecular
functions of BRCA1 and BRCA2 are still
uncertain, although they appear to be
tumor suppressor genes whose products
are involved in repairing damaged DNA.
Women who bear mutations in BRCA1 or
BRCA2 have a 36 to 85 percent lifetime
risk of developing breast cancer and a 16
to 60 percent risk of developing ovarian
cancer. Men with germ-line mutations in
BRCA2 have a 6 percent lifetime breast
cancer riska hundredfold increase over
the general male population.
BRCA1 and BRCA2 genetic tests detect
any of the over 2000 different mutations
that are known to occur within the coding
regions of these genes, but the tests have
AND
limitations. They do not detect mutations
in regulatory regions outside the coding
regionmutations that could cause aber-
rant expression of these genes. Also, little
is known about how any particular muta-
tion manifests itself in terms of cancer
risk, and the effects of each mutation may
be modified by environmental factors and
by interactions with other susceptibility
genes.
Many patients at risk for familial breast
cancer opt to undergo genetic testing.
These patients feel that test results will
help them to prevent breast or ovarian
cancers, will guide them in childbearing
decisions, and allow them to inform fami-
ly members at risk. But none of these ben-
efits is clear-cut.
A woman whose BRCA test results are
negative may be relieved and feel that she
is not subject to familial breast cancer.
However, her risk of developing breast
cancer is still 12 percent (the population
risk), and she should continue to monitor
for the disease. Also, a negative BRCA
genetic test does not eliminate the possi-
bility that she bears an inherited mutation
in another gene that increases breast can-
cer risk or that BRCA1 or BRCA2 gene
mutations exist in regions of the genes
that are inaccessible to current genetic
tests.
A woman whose test results are positive
faces difficult choices. Her treatment
options are poor, consisting of close
monitoring, prophylactic mastectomy, or
oophorectomy (removal of breasts and
ovaries respectively) and taking drugs
such as tamoxifen. Prophylactic surgery
reduces her risks but does not eliminate
them, as cancers can still occur in tissues
that remain after surgery. Drugs such as
tamoxifen reduce her risks but have seri-
ous side effects. Genetic tests affect not
only the patient but also the patients
entire family. People often experience
fear, anxiety, and guilt on learning that
they are carriers of a genetic disease.
Studies show that people who refuse
genetic test results often suffer from even
more anxiety than those who opt to learn
the results. Confidentiality is also a major
concern. Patients fear that their genetic
test results may be leaked to insurance
companies or employers, jeopardizing
their prospects for jobs or affordable
Genetics, Technology, and Society 371
SOCIETY
health and life insurance. One study
shows that a quarter of eligible patients
refuse BRCA gene testing because of con-
cerns about cost, confidentiality, and
potential discrimination.
Genetic testing is such a new develop-
ment that the health system has lagged
behind the science. Because genetic test-
ing has both psychological and medical
ambiguities, genetic counseling is impera-
tive for patients and their families. How-
ever, there are insufficient numbers of
genetic counselors with experience in
genetic testing, and even in the most qual-
ified hands, issues are complex and
difficult. Physicians often have limited
knowledge of human clinical genetics and
feel inadequate to advise their patients.
The federal government and the insurance
industries have yet to develop comprehen-
sive policies concerning genetic tests and
genetic information. Given the unclear
interpretation of BRCA genetic tests, the
relatively ineffective treatment options,
and the potential for psychological and
societal side effects, it is not surprising
that only about 60 percent of familial
breast cancer patients and their families
decide to take the genetic tests.
The unanswered questions about
BRCA1 and BRCA2 genetic testing are
many and important. What cancer risks
are associated with which mutations?
Should all people have access to BRCA
tests or only those at high risk? How can
we ensure that the high costs of genetic
tests and counseling do not limit this new
technology to only a portion of the popu-
lation? As we develop genetic tests for
more and more diseases over the next few
decades, our struggle with these issues
will continue to grow.
Reference
Surbone, A. 2001. Ethical implications of
genetic testing for breast cancer sus-
ceptibility. Crit. Rev. in Onc./Hem. 40:
149157.
Web Sites
Genetic Testing for BRCA1 and BRCA2:
Its Your Choice [online]. National
Institutes of Health, 2002. http://
cis.nci.nih.gov/fact/3_62.htm
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 372

372 Chapter 16 Cell-Cycle Regulation and Cancer

CHAPTER SUMMARY
1. Cancer is a genetic disease, predominantly of somatic cells.
About 1 percent of cancers are associated with germ-line muta-
tions that increase the susceptibility to certain cancers.
2. Cancer cells show two basic properties: abnormal cell prolifera-
tion and a propensity to spread and invade other parts of the
body (metastasis). Genes controlling these aspects of cellular
function are either mutated or expressed inappropriately in can-
cer cells.
3. Cancers are clonal, meaning that all cells within a tumor orig-
inate from a single cell that contains a number of mutations.
4. The development of cancer is a multistep process, requiring mu-
tations in several cancer-related genes.
5. Cancer cells show high rates of mutation, chromosomal abnor-
malities, and genomic instability. This leads to the accumulation
of mutations in specific genes that control aspects of cell prolif-
eration, apoptosis, differentiation, DNA repair, cell migration
and cell-cell contact.
6. Cancer cells have defects in the regulation of cell-cycle pro-
gression, cell-cycle checkpoints, and signal transduction
pathways.
7. Proto-oncogenes are normal genes that promote cell growth and
division. When proto-oncogenes are mutated or mis-expressed
in cancer cells, they are known as oncogenes.
8. Tumor suppressor genes normally regulate cell-cycle check-
points and apoptosis. When tumor suppressor genes are mutated
or inactivated, cells cannot correct DNA damage. This leads to
accumulation of mutations that may cause cancer.
9. Inherited mutations in cancer-susceptibility genes are not suffi-
cient to trigger cancer. A second somatic mutation, in the other
copy of the gene, is necessary to trigger tumorigenesis. In addi-
tion, mutations in other cancer-related genes are necessary for
the development of hereditary cancers.
10. RNA and DNA tumor viruses contribute to cancers by stimulat-
ing cells to proliferate, introducing new oncogenes, interfering
with the cells normal tumor suppressor gene products, or stimu-
lating the expression of a cells proto-oncogenes.
11. Environmental agents such as chemicals, radiation, viruses, and
chronic infections contribute to the development of cancer. The
most significant environmental factors that affect human cancers
are tobacco smoke, diet, and natural radiation.

KEY TERMS
acute transforming retroviruses, 369 G2/M checkpoint, 362 protein kinase, 360
aflatoxin, 370 hepatitis B virus, 369 proto-oncogenes, 364
apoptosis, 363 hereditary nonpolyposis colorectal cancer provirus, 369
basal lamina, 367 (HNPCC), 361 programmed cell death, 363
benign tumor, 359 human T-cell leukemia virus ras gene family, 365
Burkitts lymphoma, 359 (HTLV-1), 369 RB1 (retinoblastoma 1), 366
carcinogens, 360 loss of heterozygosity, 368 retinoblastoma, 366
caspases, 363 M checkpoint, 362 retinoblastoma protein (pRB), 366
cell adhesion molecules, 367 malignant tumors, 359 retroviruses, 369
cell proliferation, 359 metalloproteinases, 367 reverse transcriptase, 369
chronic myelogenous leukemia metastases, 359 Rous sarcoma virus (RSV), 369
(CML), 360 metastasis, 359 signal transduction, 362
cyclin-dependent kinases (CDKs), 362 mutator phenotype, 360 tissue inhibitors of metalloproteinases
cyclins, 362 oncogene, 364 (TIMPs), 367
Epstein-Barr virus, 369 papilloma viruses (HPV 16, 18), 369 transformed, 368
extracellular matrix, 367 p53 gene, 365 tumorigenesis, 360
familial adenomatous polyposis (FAP), 368 Philadelphia chromosome, 360 tumor suppressor genes, 364
G1/S checkpoint, 362 polyp, 368

AU: this term is not bold in chapter text. Pls advise

KLUGMC16_357-375v2 9/1/06 2:15 PM Page 373
INSIGHTS AND SOLUTIONS
1. In disorders such as retinoblastoma, a mutation in one allele of
the RB1 gene can be inherited from the germ line, causing an
autosomal dominant predisposition to the development of eye
tumors. To develop tumors, a somatic mutation in the second
copy of the RB1 gene is necessary, indicating that the mutation
itself acts as a recessive trait. Given that the first mutation can be
inherited, in what ways can a second mutational event occur?
Solution: In considering how this second mutation arises, we
must look at several types of mutational events, including
changes in nucleotide sequence and events that involve whole
chromosomes or chromosome parts. Retinoblastoma results
when both copies of the RB1 locus are lost or inactivated. With
this in mind, you must first list the phenomena that can result in a
mutational loss or the inactivation of a gene.
One way the second RB1 mutation can occur is by a nucleotide
alteration that converts the remaining normal RB1 allele to a
mutant form. This alteration can occur through a nucleotide sub-
stitution or by a frameshift mutation caused by the insertion or
deletion of nucleotides during replication. A second mechanism
involves the loss of the chromosome carrying the normal allele.
This event would take place during mitosis, resulting in chromo-
some 13 monosomy and leaving the mutant copy of the gene as
the only RB1 allele. This mechanism does not necessarily involve
loss of the entire chromosome; deletion of the long arm (RB1 is
on 13q) or an interstitial deletion involving the RB1 locus and
some surrounding material would have the same result. Alterna-
tively, a chromosome aberration involving loss of the normal
copy of the RB1 gene might be followed by duplication of the
chromosome carrying the mutant allele. Two copies of chromo-
some 13 would be restored to the cell, but the normal RB1 allele
would not be present. Finally, a recombination event followed by
chromosome segregation could produce a homozygous combina-
tion of mutant RB1 alleles.
2. Proto-oncogenes can be converted to oncogenes in a number of
different ways. In some cases, the proto-oncogene itself becomes
amplified up to hundreds of times in a cancer cell. An example is
the cyclin D1 gene, which is amplified in some cancers. In other
cases, the proto-oncogene may be mutated in a limited number of
specific ways leading to alterations in the gene products struc-
ture. The ras gene is an example of a proto-oncogene that
becomes oncogenic after suffering point mutations in specific
regions of the gene. Explain why these two proto-oncogenes
(cyclin D1 and ras) undergo such different alterations in order to
convert them into oncogenes.
Solution: The first step to solving this question is to understand
the normal functions of these proto-oncogenes and to think about
how either amplification or mutation would affect each of these
functions.
The cyclin D1 protein regulates progression of the cell cycle
from G1 into S phase, by binding to CDK4 and activating this
kinase. The cyclin D1/CDK4 complex phosphorylates a number
of proteins including pRB, which in turn activates other proteins
in a cascade that results in transcription of genes whose products
are necessary for DNA replication in S phase. The simplest way
to increase the activity of cyclin D1 would be to increase the
Insights and Solutions 373
number of cyclin D1 molecules available for binding to the cells
endogenous CDK4 molecules. This can be accomplished by sev-
eral mechanisms, including amplification of the cyclin D1 gene.
In contrast, a point mutation in the cyclin D1 gene would most
likely interfere with the ability of the cyclin D1 protein to bind to
CDK4; hence, mutations within the gene would probably repress
cell-cycle progression rather than stimulate it.
The ras gene product is a signal transduction protein that oper-
ates as an on/off switch in response to external stimulation by
growth factors. It does so by binding either GTP (the on state) or
GDP (the off state). Oncogenic mutations in the ras gene occur in
specific regions that alter the ability of the Ras protein to
exchange GDP for GTP. Oncogenic Ras proteins are locked in
the on conformation, bound to GTP. In this way, they constantly
stimulate the cell to divide. An amplification of the ras gene
would simply provide more molecules of normal Ras protein,
which would still be capable of on/off regulation. Hence, simple
amplification of ras would be less likely to be oncogenic.
3. Explain why many oncogenic viruses contain genes whose
products interact with tumor suppressor proteins.
Solution: In order to answer this question, it is useful to consider
what viruses try to accomplish in cells and what the roles of
tumor suppressors are in normal cells.
The goal of oncogenic viruses is not to cause cancer but to max-
imize the potential for viral replication. Viruses are relatively
simple entities, consisting of only a nucleic acid genomeeither
RNA or DNAand a relatively small number of structural and
enzymatic proteins. They depend on their host cells for much of
the biosynthetic machinery and structural components necessary
to replicate their genomes and assemble new virus particles. For
example, many viruses require the host cells RNA polymerase II
enzyme, transcription factors, and ribonucleotide precursors in
order to transcribe their viral genes. They also require compo-
nents of the host cells DNA replication machinery to replicate
viral genomes. Hence, the ideal host cell for a virus infection is
one that is within the cell cycle, preferably in late G1 to early S
phase.
Most cells in a higher eukaryote such as humans are quiescent
(in G0 phase). In order to stimulate the infected cell to enter the
cell cycle and become primed for DNA replication, many viruses
contain genes that encode growth-stimulating proteins. As we
learned in this chapter, tumor suppressor proteins are those
involved in either restraining the cell cycle at checkpoints or in
triggering the process of programmed cell death. Both of these
functions are inhibitory to the goals of a typical virus; therefore,
many viruses have evolved methods to inactivate tumor suppres-
sors. One of the ways in which viral proteins can inactivate tumor
suppressor proteins is to bind to them and inhibit their functions.
The tumor suppressors p53 and pRB are common targets of viral
regulatory proteins, such as the E6 and E7 proteins of HPV 16
and 18. By inactivating tumor suppressors, these viruses are able
to maintain the cell within the cell cycle. For the host, however,
this growth stimulation in the absence of functional cell-cycle
checkpoints can lead to increased mutation accumulation and
possible tumorigenesis.
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 374
374 Chapter 16 Cell-Cycle Regulation and Cancer
PROBLEMS AND DISCUSSION QUESTIONS
1. As a genetic counselor, you are asked to assess the risk for a cou-
ple with a family history of retinoblastoma who are thinking
about having children. Both the husband and wife are phenotyp-
ically normal, but the husband has a sister with familial
retinoblastoma in both eyes. What is the probability that this
couple will have a child with retinoblastoma? Are there any tests
that you could recommend to help in this assessment?
2. What events occur in each phase of the cell cycle? Which phase
is most variable in length?
3. Where are the major regulatory points in the cell cycle?
4. List the functions of kinases and cyclins, and describe how they
interact to cause cells to move through the cell cycle.
5. (a) How does pRB function to keep cells at the G1 checkpoint?
(b) How do cells get past the G1 checkpoint to move into S
phase?
6. What is the difference between saying that cancer is inherited
and saying that the predisposition to cancer is inherited?
7. Although tobacco smoking is responsible for a large number of
human cancers, not all smokers develop cancer. Similarly, some
people who inherit mutations in the tumor suppressor genes p53
or RB1 never develop cancer. Describe some reasons for these
observations.
8. What is apoptosis, and under what circumstances do cells under-
go this process?
9. Define tumor suppressor genes. Why is a mutation in a single
copy of a tumor suppressor gene expected to behave as a reces-
sive gene?
10. In the Rous sarcoma virus (RSV) genome, the host cell proto-
oncogene is converted into an oncogene. How does this conver-
sion occur?
11. Part of the Ras protein is associated with the plasma membrane,
and part extends into the cytoplasm. How does the Ras protein
transmit a signal from outside the cell into the cytoplasm? What
happens in cases where the ras gene is mutated?
12. If a cell suffers damage to its DNA while in S phase, how can
this damage be repaired before the cell enters mitosis?
13. Distinguish between oncogenes and proto-oncogenes. In what
ways can proto-oncogenes be converted to oncogenes?
14. Of the two classes of genes associated with cancer, tumor sup-
pressor genes and oncogenes, mutations in which group can be
considered gain-of-function mutations? In which group are the
loss-of-function mutations? Explain.
15. How do translocations such as the Philadelphia chromosome
contribute to cancer?
16. Given that cancers can be environmentally induced and that
some environmental factors are the result of lifestyle choices
such as smoking, sun exposure, and diet, what percentage of the
money spent on cancer research do you think should be devoted
to research and education on preventing cancer rather than on
finding cancer cures?
17. In CML, leukemic blood cells can be distinguished from other
cells of the body by the presence of a functional BCR-ABL hy-
brid protein. Explain how this characteristic provides an oppor-
tunity to develop a therapeutic approach to a treatment for CML.
18. How do normal cells protect themselves from accumulating mu-
tations in genes that could lead to cancer? How do cancer cells
differ from normal cells in these processes?
19. Describe the difference between an acute transforming virus and
a virus that does not cause tumors.
20. Explain how environmental agents such as chemicals and radia-
tion cause cancer.
21. Radiotherapy (treatment with ionizing radiation) is one of the
most effective current cancer treatments. It works by damaging
DNA and other cellular components. In which ways could radio-
therapy control or cure cancer, and why does radiotherapy often
have significant side effects?
22. Assume that a young woman in a suspected breast cancer family
takes the BRCA1 and BRCA2 genetic tests and receives negative
results. That is, she does not test positive for the mutant alleles of
BRCA1 or BRCA2. Can she consider herself free of risk for
breast cancer?
23. People with a genetic condition known as Li-Fraumeni syn-
drome inherit one mutant copy of the p53 gene. These people
have a high risk of developing a number of different cancers,
such as breast cancer, leukemias, bone cancers, adrenocortical
tumors, and brain tumors. Explain how mutations in one cancer-
related gene can give rise to such a diverse range of tumors.
24. Explain the differences between a benign and malignant tumor.
25. As part of a cancer research project, you have discovered a gene
that is mutated in many metastatic tumors. After determining the
DNA sequence of this gene, you compare the sequence with
those of other genes in the human genome sequence database.
Your gene appears to code for an amino acid sequence that re-
sembles sequences found in some serine proteases. Conjecture
how your new gene might contribute to the development of high-
ly invasive cancers.
26. A study by Bose and colleagues (1998. Blood 92: 33623367)
and a previous study by Biernaux and others (1996. Bone Mar-
row Transplant 17: (Suppl. 3) S45S47) showed that BCR-ABL
fusion gene transcripts can be detected in 25 to 30 percent of
healthy adults who do not develop chronic myelogenous
leukemia (CML). Explain how these individuals can carry a fu-
sion gene that is transcriptionally active and yet do not develop
CML.
27. Those who inherit a mutant allele of the RB1 gene are at risk for
developing a bone cancer called osteosarcoma. You suspect that
in these cases, osteosarcoma requires a mutation in the second
RB1 allele and have cultured some osteosarcoma cells and
obtained a cDNA clone of a normal human RB1 gene. A
colleague sends you a research paper revealing that a strain of
cancer-prone mice develop malignant tumors when injected
with osteosarcoma cells, and you obtain these mice. Using these
three resources, what experiments would you perform to deter-
mine (a) whether osteosarcoma cells carry two RB1 mutations,
(b) whether osteosarcoma cells produce any pRB protein, and
(c) if the addition of a normal RB1 gene will change the cancer-
causing potential of osteosarcoma cells?
28. The compound benzo[a]pyrene is found in cigarette smoke. This
compound chemically modifies guanine bases in DNA. Such ab-
normal bases are usually removed by an enzyme that hydrolyzes
the base, leaving an apurinic site. If such a site is left unrepaired,
an adenine is preferentially inserted across from the apurinic
site. In a study of lung cancer patients (Harris, A. 1991. Nature
350: 377378), tumor cells from 15 out of 25 patients had a G to
T transversion in the p53 gene, which has a known role in cancer
formation. You are testifying as an expert witness in a court case
in which the widow of a man who was a lifelong smoker and
died of lung cancer is suing a company for manufacturing the
KLUGMC16_357-375v2 9/1/06 2:15 PM Page 375

Problems and Discussion Questions 375

TABLE 16.5 Predisposing Mutations in BRCA1

Mutation

Kindred Codon Nucleotide Change Coding Effect Frequency in Control

Chromosomes

1901 24 -11 bp Frameshift or splice 0/180

2082 1313 C:T Gln : Stop 0/170
1910 1756 Extra C Frameshift 0/162
2099 1775 T:G Met : Arg 0/120
2035 NA* ? Loss of transcript NA*
Source: 1994. Science 266: 6671.

*NA indicates not applicable, as the regulatory mutation is inferred, and the position has not been identified.

tobacco products that killed her husband. What would you tell
the jury? (Science only pleaseno personal expositions on
lawyers or the legal system!)
29. Table 16.5 summarizes some of the data that have been collected
on BRCA1 mutations in families with a high incidence of both
early-onset breast and ovarian cancer. Table 16.6 shows neutral
polymorphisms found in control families (with no increased fre-
quency of breast and ovarian cancer). (a) Note the coding effect
of the mutation found in kindred group 2082 in Table 16.5. This
results from a single base-pair substitution. Draw the normal dou-
ble-stranded DNA sequence for this codon (with the 5 and 3
ends labeled), and show the sequence of events that generated
this mutation, assuming that it resulted from an uncorrected mis-
match event during DNA replication. (b) Examine the types of
mutations that are listed in Table 16.5 and determine if the
BRCA1 gene is likely to be a tumor suppressor gene or an onco-
gene. (c) Although the mutations in Table 16.5 are clearly delete-
rious and cause breast cancer in women at very young ages, each
of the kindred groups had at least one woman who carried the
mutation but lived until age 80 without developing cancer. Name
at least two different mechanisms (or variables) that could under-
lie variation in the expression of a mutant phenotype and propose
an explanation for the incomplete penetrance of this mutation.
How do these mechanisms or variables relate to this explanation?
30. Examine Table 16.6. (a) What is meant by a neutral polymor-
phism? (b) What is the significance of this table in the context of
examining a family or population for BRCA1 mutations that pre-
dispose an individual to cancer? (c) Is the PM2 polymorphism
likely to result in a neutral missense mutation or a silent muta-
tion? (d) Answer part (c) for the PM3 polymorphism.

TABLE 16.6 Neutral Polymorphisms in BRCA1

Frequency in Control Chromosomes*

Name Codon Base in

Location Codon A C G T

PM1 317 2 152 0 10 0

PM6 878 2 0 55 0 100

PM7 1190 2 109 0 53 0

PM2 1443 3 0 115 0 58

PM3 1619 1 116 0 52 0

*The number of chromosomes with a particular base at the indicated polymorphic site (A, C, G, or T) is shown.